DISCUSSION: It must be remarked that no language in which CVC roots predominate is likely to squander a CVC-sequence on something as specialized as a 'mulberry' — for economy of expression, more commonly used concepts will occupy any given sequence.

And so it is with Nostratic *mor- which, we hope to demonstrate, meant 'black'.

*MO meant, primarily, 'blood'. 'Blood' has two characteristic colors: 'red' when fresh, 'black' when not. In derivations from Nostratic *mo, we see both meanings in various derivatives: e.g. in Egyptian mrS (Coptic [S] morS), 'be (light) red' (cf. IE *merg^w-, 'dark'); and B merre, 'red (of hair)'. On the other hand, we have mr(-)wr.yt, 'black stork'. This element is probably also present in mrH, 'decay', i.e. 'darken with age', corresponding to IE *merk- (for **merkh-), 'decay'. In addition to *moro-, 'blackberry, mulberry', IE has another root which unlocks the key to the reconstruction of the root: (3.) (mer-), mor(-u)-, 'blacken, dark color, dirt-spot'.

We would have expected the *RHA, 'color', to have been used here (**MO-RHA) but Egyptian r can only represent Nostratic *r(h)o). However, it is possibly that the earliest form was PL *MO-*RHA-*FA-, 'blood-color-flatly-circular' = 'black spot', Nostratic *mora:w-, which was resolved into **moro:- in pre-Egyptian. Nostratic *mora:w- would relate to the IE attested form *moru-, 'dirt-spot'.

What decides us in favor of this interpretation is Sumerian muru₉, 'fog, mist, haze'; and muru, a reading for a sign that means 'spotty, be unclean, poor repair', or 'dejection', which are echoed in Armenian mr^.ayl, 'darkness, fog, cloud'; and IE *merk-, 'morose'. Accordingly, we will regard Nostratic mora:- as primary, with a frequent extension of *-w.

This root seems to have been particularly well-represented in Dravidian, where we have *mur-, 'dry up, wither (darken), burn, singe, scorch (darken), fuel'.

We, therefore, reconstruct PL *MO-*RHA-(*FA-), 'blood-color(-flatly-circular)' = 'be black or red (black spot)'.

Rédei has Uralic *mura- for 'rubus chamaemorus', 'cloudberry', which is in the blackberry family, and a large selection of other berries, which all seem to be red or black.

No suitable Altaic cognate may be presently identified.

Accordingly, we amend:

(532)PN *mora:(-w)- "to be black; to be red (black spot)" > PIE (3.) *(mer-) "to blacken; dark color"; *moru- "dirt-spot"; *moro- "blackberry, mulberry (black/red one)"; PAA *mar(aw)- in Egyptian mrS (Coptic [S] morS, 'to be (light) red'); probably in mr(-)wr.yt "black stork"; in B merre "red (of hair)"; PU *mura- (for **mura:) "rubus chamaemorus (cloudberry), blackberry[?]"; PD *mur- (Starostin) "to dry up, to wither (to darken), to burn, to singe, to scorch (to blacken); fuel; S muru (for **murû) "spotty; to be unclean; poor repair; dejection"[?] (source: PL *MO-RHA-[FA-])

DISCUSSION: Orel-Stobova, to their credit, do identify the simplest Hamito-Semitic form of a very important root: "*ma-/*mi- "mouth"", which, however, is inaccurate in two ways: 1) the meaning is actually 'tongue', representing Nostratic *me and PL *ME; and, if the vowel quality had been preserved in Hamito-Semitic, it would have been *i only.

We find this element in IE only in combination: PL *ME-*FA, 'tongue-flatly-circular' = 'mouth', seen in IE (1.) *mu/u:-, 'mouth'. Sumerian mu₂ means 'sing', and mu₇, 'shout', and 'incantation', which compares with other meanings of IE (1.) *mu/u:-: 'cry out, speak incomprehensibly'. The form *mu:- indicates a medial 'laryngeal' but in the absence of a (Hamito-)Semitic cognate, it is impossible to specify it.

Surprisingly, the simplex occurs in IE in a slightly different setting. Bomhard has it as:

(524)PN *mi-/*me- interrogative pronoun stem, *ma-/*m6- relative pronoun stem > PIE *me-/*mo- interrogative and relative pronoun stem; PK *mi-n- interrogative pronoun; PAA *ma-/*m6- interrogative and relative pronoun stem; PU *mi- interrogative and relative pronoun stem; Altaic: Turkish mi, mI, mu, mü interrogative particle; S me-na-àm "when?"; me-a "where?"; me-šè "where to?"; PE (Proto-Eskimo) enclitic particle *mi "what about?"

In addition to 'tongue', PL *ME also meant 'call out'. The use of it as an interrogative is just our interpretation of what is actually an imperative: 'call out he goes' = 'who goes?'; 'call out price' = 'what is the price?'

Bomhard lists IE "*me-/*mo- interrogative and relative pronoun stem" but this is a root unknown to Pokorny. In another publication, he attempts to substantiate this proposed IE root with examples like Breton ma, 'that'. It would be better if we had other IE examples but this might be acceptable as a reflex of Nostratic *ma, 'where' or even Nostratic *me, 'what'. He also cites Tocharian B mäksu, 'which?', and mäkte, 'how?' I think it likelier that these are derivatives of *mäk, '?'; and they will not be included in the amended listing. In addition, he lists Tocharian A mänt/mät, 'how'. Unless the terminations can be convincingly analyzed, I cannot accept these as cognates either. Finally, he lists Hittite maši,'how much?' I find this similarly unconvincing. Accordingly, only the suggested Breton cognate will be listed. It is likely that IE had and lost this root, substituting *k^we-/*k^wo-, 'who?, what?', from Nostratic *xha- (PL *XHA, 'large quantity of animate beings').

Sumerian has mu, 'hero' (not listed in John Halloran's Sumerian Lexicon); and mua/muia/mu₆, 'hero, man'. I assume that the meaning 'hero' reflects Nostratic *mo?, i.e. PL *MHO-*?A-, '(hu)man'+stative='(be-)manly'='hero(ic)', and reflects the root identified by Orel-Stobova as *mu?-, 'man'. The Sumerian form (mu) should probably be amended to **mû to represent the lengthening due to the loss of the glottal stop. On the other hand, mu₆ probably represents the simplex: '(hu)man'. We simply do not know if Nostratic *o: normally led to a long vowel in Sumerian since length is unindicated; however, the existence of mua for 'hero' suggests that the final -a did not simply lengthen an already long vowel but had to remain in a separate syllable (**mûa) while muia probably indicates the same pronunciation with an epenthetic glide.

It is tempting to regard Egyptian mn, 'someone', as belonging here; but, as we shall see below, mn is properly **mjn; and it is not possible with the data at our disposal to decide whether it should properly be reconstructed as Nostratic *mo:yan-, our amendment of Orel-Stobova's Hamito-Semitic *mayan-, 'man', or not but this is somewhat probable; this would result in Egyptian mjn.

Rather than interpreting Sumerian me-a, 'where?', as a combination of me, 'what', plus locative -a, I think it likelier that the a is a mater lectionis for the reading ma₆, another reading of the sign read me. I would therefore favor amending me-še₃, 'whither?' to ma₆-še₃, 'whither?'.

In the case of Sumerian me-na-am₃, 'when?', I believe we have the root seen in Nostratic *me. The form we expect in Sumerian is **mi, and me also reads mi₃, to which we will amend the reading. Sumerian am₃ is recognized as an enclitic copular form: 'it is'. So, we have a phrase meaning 'when?' with the elements 'what . . . is it?' Obviously, the meaning of na in this context must be 'time'. There is no Nostratic *na meaning 'time' but a PL root for 'time' is reconstructable as *HHE-*NO-, 'move-up-from-jar'='clepsydra (with a float time-indicator)'='time'. This root is found in Egyptian nw (for **jnw), 'time'; IE (2.) *en-, 'year'; and Sumerian en_2,3 (for **in_x), 'time', which also means 'when'. It appears that the phrase was originally mi₃-in_x-am₃, in which the second i coalesced with the first, and in which the n attached itself to the following word: [mi:-nam].

No suitable Dravidian cognate can be found at present fr any of these forms.

For some reason, Bomhard has included "*ma-/*m6- relative pronoun stem" with the interrogative element in #524, for which there is no good reason that I can see. There are actually two relative pronoun stems reconstructable for Nostratic: *ma, 'mound', 'where', which refers back to nouns of place; and *mo:, 'human', 'who', which refers back to persons. We will abstract these from his entry;

and we will amend to include:

(Addition AC)PN *ma- "where" (relative pronoun) > PIE *me-/*mo- "where[?]" in Breton ma "what"; S me-a (for **ma₆[?]) "where?"; me-še₃ (for **ma₆-še₃) "whither?" (source: PL *MA-)

and

(Addition AD)PN *mo:- "human, who" (personal relative pronoun) > PIE only as *mo-n-u-s (for **mo:-n-u-s) "(hu)man"; PAA in *ma-n- '"who?"; Altaic: Turkish mu, mü "who?"; S only as mu₆ (possibly **mû) "(hu)man" (source: PL *MHO-)

and finally amend:

(524)PN *me- "call out; what?" (interrogative construction) > PIE *me-/*mo- "what[?]" in Breton ma "what"; PAA *ma "what?"; PK in *mi-n- interrogative pronoun "who?"; PU *mi- interrogative and relative pronoun stem; Altaic: Turkish mi, mI "what?"; S in me-na-am₃ (for **mi₃-in_x-am₃) "what time is it?, when?"; PE (Proto-Eskimo) *mi "what about?" (enclitic particle) (source: PL *ME-)

DISCUSSION: There is yet another important derivation of Nostratic *me (PL *ME) that needs to be investigated. It is represented in Bomhard's dictionary as:

This is another productive use of Nostratic *me, which has the secondary connotation of 'call out'. In this particular case, it is functioning as an equivalent to 'speaker', the function of the first person singular of pronouns.

The amendments we will make to this entry are minimal.

The word is not attested in Semitic languages but, in another publication, Bomhard does provide some Chadic and other African cognates that appear to be plausible for establishing its presence in PAA.

In the main dialect of Sumerian (Emeku) as well as in Emesal, a 1st person plural suffix me, which we prefer to read mi₃ (another reading of the same sign), exists. The normal Emesal form for the ergative case is me-e, where -e is the ergative suffix. Here again we would prefer to read mi₃. The form cited by Bomhard, ma, is fairly likely to be a contraction of mi₃-a, and probably is **mâ. Sumerian -a is normally the sign of the locative case; and mi₃-a may probably be regarded as a substitution for mi₃-e or possibly just a dissimilation of it.

So we will amend to:

DISCUSSION: There is another root that must be discussed in this context:

This is a rather surprising entry: what are we to believe the Nostratic root is — *ma-/*m6- or *mu-/*mo-? In my opinion, based on the forms cited, there is no reason to reconstruct *ma-/*m6-. Nostratic (*mu-/)*mo-, which I would amend to mo:- can explain the alleged cognates detailed. At first glance, these appear to be derivatives of Nostratic *mo:-, '(hu)man', listed above in Addition AD, used as an indefinite pronoun rather than as a relative pronoun.

I am including Bomhard's PIE *mo- even though I have no idea of the circumstances which led him to reconstruct it.

Accordingly, we amend:

After seeing some of Bomhard's earlier work, I suggested to him a different method of organizing the entries in his dictionary; he declined. Let me now insist that any method of organization that has one entry *mi-/*me- as #524 and another *mi-/*me- as #540 is in serious need of rethinking. There is no traditionally transmitted alphabet-order for Nostratic; and for most proto-languages, proto-linguists have seen fit to utilize simple alphabetic order. I have utilized it in the alphabetic list of roots investigated in this essay with no problems at all. I strongly advise Bomhard in any possible future efforts to do the same. It is, as a result of his idiosyncratic listing-order, extremely difficult to find any entry easily in his dictionary; this is an encumbrance without any possible justification.

Another monosyllabic root is found in Bomhard is:

This is one of Bomhard's poorest entries. The only way a 'laryngeal' can be determined is if it occurs in an Afrasian cognate. For the prohibitive particle, Arabic shows mâ not *ma?; and reconstruction of the glottal stop is unjustified.

IE *me: indicates that the original Nostratic vowel was *e. Since any 'laryngeal' could lengthen *e to e:, the choice of the glottal stop is arbitrary and unjustified by the data. Secondly, IE *e: cannot be the result of a lengthened Nostratic *a, which would result only in IE *a:.

With lengthening but the display of no laryngal or pharyngal in Semitic, the inescapable conclusion must be that the length in both Arabic and IE is due to an aspirated nasal (PL MH) that has lengthened the following vowel (E) when it was lost in Nostratic; this leads us to a reconstruction of Nostratic *me:, reflecting PL *MHE-, 'come off of'='cease planning to . . .'.

Although I include Bomhard's proposed Kartvelian cognates, I have no easy way to verify that Nostratic *e can result in Svan a:, apparently the datum on which Bomhard relied for his Nostratic reconstruction of *a. However, the IE evidence is unequivocal: the Nostratic vowel must have been *e.

Accordingly, we amend:

DISCUSSION: We will next discuss another root of the similar form:

Anyone writing in 1996 who did not know that Egyptian 3 in Old and Middle Egyptian represents a kind of [r] should be faulted for not keeping up with developments; this has been as firmly established as these matters can be for some time. I identified Egyptian 3 as [r] independently in 1975 but it took many years after Carleton Hodge subsequently proposed it for the idea to become accepted among Egyptologists.

IE *me: indicates that the original Nostratic vowel was *e. The hieroglyph for writing it pictures a 'sickle', an implement which we know was fitted with small triangular microliths in some of its earliest manifestations.

This leads us to a reconstruction of Nostratic *me?-, reflecting PL *ME-*?A, 'tongue'+stative='tongued'='saw-toothed'='sickle'.

That Egyptian m3 is to be amended to **mj3 is strongly suggested by Coptic mie:, 'lioness', written in ME as m3-jr.t. In addition, Egyptian m3(3) (for **mj3), 'see', is cognate with IE (s)meir- in Latin mi:r-or, 'look at in admiration'.

The final Egyptian -3 is Nostratic *-r^y, PL *-RE, 'apply'; and in this context: 'apply the sickle=reap, mow'. Apparently, this formant was not retained or adopted by IE, which uses a *-t-formant to the same purpose.

There is a possible Sumerian cognate in me₃ (for **mî_x), 'battle', which would be a similar development to what we see, e.g. in Modern German , 'slaughter', derived from IE *me(:)t- mentioned above.

Accordingly, we amend:

DISCUSSION: We will next discuss another root of the similar form, listed by Bomhard inconveniently as:

(408)PN *miH/*meH "to measure, to mark off" > PIE *miHh- [*meHh-] (> *me:) "to measure, to mark off"; Afroasiatic: Egyptian mH "cubit, forearm"

We have demonstrated throughout this critique that Egyptian H derives from Nostratic *k?x and *khx before *a or *e (before Nostratic *o, both are Egyptian x (hook-h). Any Nostratic laryngal or pharyngal appears as Egyptian j if before *a or *e, as h if before *o.

Bomhard's reconstruction of a phoneme Hh for earliest Indo-European might be true but it is totally irrelevant since it has no observable effect on the forms into which IE will develop, i.e., any Nostratic laryngal or pharyngal will lengthen an earliest IE vowel, and prevent it from developing into the *e/*o-Ablaut.

As we have stated repeatedly above, without a Semitic cognate, it is impossible to precisely determine the nature of a Nostratic laryngal or pharyngal, appearing in IE as a lengthening of the vowel. Though the traces of Nostratic *hh are tantalizingly sparse for this root, I believe they may be found in Arabic maHana, 'prove something, examine a student'; and ?imtaHana, 'examine a word'; both of which suggest 'measurement'. Adding some weight to this proposal is that IE (3.) me:- has an early extension in *-n-: *me:-no-, 'measure'.

Another IE extension, *-t-, leads to *me:-ti-, 'measurement'; and *me:-to-, 'year'. This is the basis on which we would like to find the Egyptian cognate, which we believe is **mj', 'true, real'. The double asterisks indicate my amendation of what is traditionally transcribed by Egyptologists as m3'. Though this word is written with the 'sickle', it is normally not written with the 'vulture' ([r]) as is Egyptian m33, discussed above. Nostratic *me?- or *mehh- would both result in Egyptian **mj- and IE *me:-. I believe the 'sickle' properly represents only mj, and that the vulture is necessary to add the 3 in a full spelling. In addition, **mj' is written with a sign which Gardiner classifies as "doubtful" but could very well be a 'measuring rod' (see especially its OK form). This would, if true, reinforce the idea that the semantic content of 'true' was based on 'measurement', i.e. 'verification'.

Since IE *t(h) can represent Nostratic *th or *ths. Egyptian ', on the other hand, can only be Nostratic *t?so or thso. If the IE word seen as *me(:)t- and Egyptian *mj' are cognate, as appears likely to me, the Nostratic final element can only be *thso, probably representing *-THSO, 'stretch-around', which probably highlights the actual process of measurement, 'measuring'. The Egyptian word for 'truth', m3'.t which we would amend to mj'.t, is seen in Coptic (S) as me, the vowel of which is some an indication of the medial j (ME [ai] > Coptic [e]).

Knowing as we do now that IE *d represents Nostratic *t?, and that Nostratic *t?o becomes Arabic D (dotted d), we can also notice Arabic ?amHaDa(m-H-D), 'be truthful', which corresponds, in my opinion, to IE (1.) *me(:)d-, 'measure'. Based on Egyptian mt-y, 'straightforward, precise', we can probably narrow this formant down to *-T?O, 'lump', so that Nostratic *mehhet?^w means something like 'measured amount'. Egyptian mty is written with a 'phallus', the same sign used to write **mjt in entry #537.

Coptic (S) me and (B) me:i, 'truth' (Egyptian m3'.t), assure us that we are correct in amending m3'.t to **mj'.t,

A very important concept in Sumerian religion is me, the sign also reading mi₃, which means something like 'ideal norm'. I think that most readers would agree that the idea of measurement is quite close to a 'norm'; and I believe that, as mi₃, it is derived from Nostratic *mehh^y- also.

This widely distributed root is also found in Altaic *miá(-)là-, 'measure; measure', which corresponds to IE *me:l- in Old Icelandic moe:la, 'measure'. The Nostratic *-l is a reflex of PL *NHA, an ingressive formant: Nostratic *mehhel-, 'start to measure'.

We therefore reconstruct the root as consisting of PL *ME-*HHE, 'call-out-move-up-from'='announce (measurements) after inspection)' .

Accordingly, we amend:

DISCUSSION: As we saw above in entry #516, the triangular blades which were used for cutting by speakers of the Proto-Language were characterized as 'tongues' rather than 'teeth' as we would have it today.

This insight allows us to better understand how to analyze other roots like:

Firstly, I believe Bomhard's elucidation of this root should be considered overly broad. To have one root to express both the process and the result of an action seems unnecessarily ambiguous.

In fact, the two ideas, process and result, can be seen to be separate in Arabic marra, 'be bitter' (conceived as 'piercing') and mariha, ' be weakened by pain', hence, 'be in pain'. The basal meaning of 'pierce' is substantiated by Arabic maraqa, 'pierce through'.

From the Arabic evidence alone, we should reconstruct two roots: the first denoting the process of 'stabbing'; the second, the result, 'being stabbed', and its interpretation: 'be in pain', and by association, 'be sick'.

For the process, rather than IE *mer-, 'die', as Bomhard would have it, the appropriate IE root is (6.) mer-, 'disturb, anger', i.e., 'stab, jab painfully'.

The Egyptian cognate can probably be found in s(-)m3, 'kill', if this can be analyzed as 'cause to be stabbed', which is likely.

Dravidian *mer-, 'cause pain', is the form found at Starostin's website; and I will substitute it for the form Bomhard cites. The alveolar *r is a further indication that the Nostratic vowel-quality of the second syllable is to be reconstructed as *re/*r^y.

A Sumerian sign, which pictures a 'thorn', is read g[~]ir₂; and means 'knife, dagger, sword, thorn, scorpion'; it can also be read mir₃. Another sign reads mir means 'scorpion'. I, therefore, believe it is quite plausible that mir₃ was, at the least, another reading for the constellation of meanings currently ascribed to g[~]ir₂, which I would characterize as 'stabber'.

Both of these Sumerian words are attested in longer forms ending in - (g[~]iri₂/miri₃). As we have repeatedly seen, normally a Nostratic CVC form will reappear in Sumerian as CVC. The retention of the final vowel in Sumerian indicates the presence of a former Nostratic laryngal or pharyngal, which has, upon disappearing, lengthened the final vowel, causing it to remain manifest. I presume, on the basis of the Arabic evidence, that this formant was Nostratic *h(a); and that it transformed the process into a result or state, which was interpreted as 'be in pain'. I believe that further research may be able to identify a meaning of 'be in pain, be sick' for miri₃ though that is not presently the case.

We would expect that a Nostratic root of the form *mereh- would appear in Egyptian as **m3j. However, we find not this but Egyptian mr with the requisite semantic associations.

Egyptian mr, 'be sick, ill, be in pain', is written with a sign which Gardiner characterizes as a 'chisel' while mentioning that 'hairpins' are very similarly depicted. What this strongly suggests is that 'sickness' was visualized by the Egyptians as 'stabbing pain'. The same Egyptian sign is used to write mr, 'pyramid', which, from its very form, could be characterized as 'stabbing'.

Egyptian r is the reflex Nostratic *r(h)o, which will show up as l in Sumerian. If mr truly is cognate as it, on semantic grounds, surely appears to be, a basis of *m3 has to have been modified to mr by the addition of some formant, with which it has contracted to r. One possibility is Nostratic *-w (PL *FA, 'do repeatedly' — 'repeatedly being stabbed' would be an appropriate metaphor for 'constant, sharp pain'. This formant can be seen in IE *mR-wó-, 'dead', which has to be interpreted as 'the state resulting from repeated stabbing'. If a theoretical Nostratic mérew- became, under the influence of the stress-accent, *mer^w, it would provide the basis for Egyptian mr.

Notice in this context that 'do repeatedly' implies a perfective achievement of the goal of the action ('stab to death').

We therefore reconstruct the root as consisting of PL *ME-*RE(-*HA / -*FA), 'tongue'='blade'+'scratch'/'apply'='stab'(-animate stative='be stabbed, be in pain' / -'do repeatedly'='stab-repeatedly'='(cause to)be dead').

Accordingly, we amend:

but we must amend to include:

Firstly, in another publication, Bomhard gives only reputed Egyptian cognates for his word; if he had followed his usual practice, he would have written: "Afroasiatic: Egyptian mtt "middle", mtw "with"; Coptic me:te "middle".

A small imperfection in this scheme is that mtt is certainly not the OE or ME term for 'middle', which is Hr-jb. If it cannot be attested early, it, by itself, hardly can justify a "PAA" reconstruction. Another problem is that it seems to occur only in the phrase m mt.t n.t jb, which means 'gladly'. While this phrase could mean 'from the middle of the heart', it is, by no means certain that it does. However, it must be conceded that Coptic me:te lends some weight to the proposal. Egyptian mtw means 'with' not 'middle', and is not attested in Old Egyptian. This may be a borrowing from another language.

Fortunately, there is an Arabic root that suggests the Nostratic form: muSSa-t-un, 'choice part, pith'. Since Arabic S is derived from Nostratic *thso, the final syllable of IE *me-ta-, 'middle', should be amended to *thV.

We have been able to locate no cognate in any language that would give us a definite idea of the vowel-quality of the first vowel in the root. Accordingly, with reservations, we will choose *a but it might just as well be *e; *o seems to be ruled out on semantic grounds.

We therefore reconstruct the root as consisting of PL *MA-*THSO, 'breast'+formant of cardinal directions='middle'.

No suitable Altaic, Dravidian, Sumerian cognate can be presently found.

Accordingly, we amend:

DISCUSSION: As might be imagined, there are a number of important roots built on PL *ME, 'tongue'. There is another root, *¿E, 'voice', which also has to do with communication, but whereas it seems to mean primarily 'speak', *ME seems to be better understood as 'communicate, converse': 'speaking' as a two-participant situation.

It is therefore quite interesting that a root which developed into 'reckon, consider', seems to mean primarily 'discussion', implying a very modern approach to decision-making.

This root is detailed in Bomhard's:

We will begin with the Dravidian cognate *man.-. There is no sign of it at Starostin's website but there is a *min.-, 'murmur (also to one's self)', which seems to be the better choice of Dravidian cognate. Bomhard, displaying a strange understanding of retroflex articulation, attributes a final Dravidian retroflex *n. to a Nostratic palatalized *ñ, which makes no phonological sense at all.As I have demonstrated in the retroflex series of Arabic apical stops, retroflexion is associated with back-vowel articulation, i.e. *o in the case of Nostratic. Accordingly, we will consider the final PL component *NO, 'store', 'what is stored'='content'.

Using the new Dravidian cognate, *min.-, the vowel-quality tells us that the Nostratic first vowel was *e.

We therefore reconstruct the root as consisting of PL *ME-*NO, 'converse-store'='conversation'.

The Sumerian sign which reads me, 'speak, call', also reads men₂, which we would amend to **min_x.

The Uralic situation is quite complicated. Rédei reconstructs *m8n3- (*mon3-), 'say', but includes a number of supposed derivates that do not simply mean 'speak' but rather 'warn, curse, belittle', etc., and it is to be wondered if we have more than one root here (as opposed to Ostiak män-, 'say'). I am inclined to think that many of these may be built on Nostratic *mew-, 'mouth', or 'murmur', seen with an *-n-extension in Latvian maunu, 'shout' (Nostratic *mewan-). Sammallahti does not recognize it for Uralic or Finno-Ugric. I am inclined to reconstruct *mäna- for Uralic though I know that the form does not seem to be represented in many branches.

No suitable Altaic cognate can be presently found.

Accordingly, we amend:

DISCUSSION: The Nostratic form that Bomhard postulated for the root above, we find correct for his:

We will begin with the Dravidian cognate *man-. Since it is neither retroflex nor palatal (*ñ), it represents Nostratic *na. Accordingly, we will consider the final PL component *NA, 'thing'.

The Dravidian cognate, *man-, tells us that the Nostratic first vowel was *a.

Bomhard completely irresponsibly lists "to be firm, steadfast, established, enduring" as meanings for the Afrasian cognate. Yet, in another publication, where he cites words from individual Afrasian languages, not a single word is listed which would support this meaning. And, in fact, another Nostratic root, which we shall presently elucidate, properly is the source of this set; it is not in Bomhard's dictionary at all. And amazingly, he does the same for the IE root, equally with no justification. I can only imagine that he has extrapolated these meanings for the other Afrasian and IE roots from Egyptian mn, 'be firm, established', which also has the meanings of 'dwell, remain'. We assume that: 1) an Egyptian [man] has coalesced with [mayn]; or 2) the sign for **mjn was also incorrectly used to write mn. In any case, we will investigate this root in the next entry.

We therefore reconstruct the root as consisting of PL *MA-*NA, 'mound-thing'='dwelling place'.

Accordingly, we amend:

DISCUSSION: We will begin with the Egyptian sign that is used to write mn, 'be firmly fixed, established'. Gardiner, in his famous sign-list, says it depicts a "draught-board". Gardiner, whose discernment is usually far superior, has mightily erred here; and generations of Egyptologists have parroted his poor judgment. The sign clearly depicts 'land with palisades' as I am hoping any reader will quickly see. The range of meanings flows easily from this central insight; 'to establish' is 'to fence off with palisades'; and 'to be firmly fixed' should be understood in a psychological rather than a physical sense.

Egyptian mn,'remain', gives Coptic (S) moun by the perfectly regular process of backing and raising a after a nasal. Interestingly, the qualitative form is me:n; and it is to be wondered if the qualitative is derived from **mjn rather than mn.

We therefore reconstruct the root as consisting of PL *ME-*¿E-*NA, 'tongue-like=blade-thing'='palisade', 'be firmly fixed'.

Sumerian has, what I take to be, a cognate in men *for **mên, 'crown'. Anyone familiar with medieval crowns will realize than they were designed to symbolize the wearer's control over a walled city.

No suitable Altaic, Dravidian, nor Uralic cognate can be found at present.

Accordingly, we amend to include:

DISCUSSION: We have seen above how PL *MA ('wave') and PL *MO ('slippery' or 'dirty') could become components of words associated with 'water'. Now, we shall see that PL *ME, 'lick', could also be such a component.

Bomhard has missed the nuance of this root completely. Its primary meaning is not "water, liquid, fluid" but "to dampen" as we shall show below. Secondly, it should be stressed again that, without an Altaic, Dravidian, or Uralic cognate, Bomhard has no way at all, under his method, of determining the root vowel of this word. And it is suspect for him to pretend that he has. The reader should not think that Bomhard has relied on, e.g. Orel-Stobova for this determination since they reconstruct *ma?- . Ehret comes much closer with his entry #569: "*-m- (*-ma^v- ?) "to be wet"; *âm- "water""; but note that his *-ma^v- (?) does not mean 'water' but rather 'be wet'.

It is to be noticed that the idea of 'water' vs 'dampness' seems to be present in this root in Arabic but only after the addition of h, which probably corresponds to the expanded form of *mew- we see in Indo-European: *mew6-, and *mu/u:-. One of the functions of PL *HA is to serve as an animate indefinite plural (for speakers of the Proto-Language, whatever seemed to move of itself was animate; 'water' is animate). My suggestion for Arabic m-w-h would be to analyze it as: PL *ME-*FA-*HA, 'lick-do-repeatedly'='dampen'+animate indefinite plural='(aqueous) liquid'. It occurs to me that Bomhard has used Nostratic *a/*6 as a catchall since, according to his suspiciously overly generous rules of equivalency, almost any vowel in a derived language can be ascribed to the pair. And nowhere that I can find does he provide the slightest indication of which allophonic variant of his "pairs" will be applicable to any given language or application.

I guess it is obvious that, based on the data provided, only one solution is possible: to analyze the semantics of the root, and apply the knowledge of PL monosyllabic meanings to the context. This root appears to be easy to analyze since PL *HHA has been determined to mean 'water'; and PL *ME, 'tongue', is the natural source of 'dampness' through 'licking'.

We therefore reconstruct the root as consisting of PL *ME-*FA, 'lick-do-repeatedly'='dampen'.

No suitable Altaic, Dravidian (though *mi:-, 'wash', is probably distantly related), nor Uralic (though *mukca-, 'travel over water', may be of interest) cognate is currently ascertainable.

Accordingly, we amend:

DISCUSSION: Another root which shows up in terms pertaining to 'water' is PL *NA, 'inside'; we see it in Bomhard's

Bomhard has missed the nuance of this root completely. Its primary meaning is not "to moisten, to wet" but, as we shall show below, '(to exude) interior moisture'.

We should first notice Egyptian nt(-)nt, 'perspiration'; and nt.t, 'outflow'; then we may properly understand nt, 'water, flood-water'; this terminology is a result of the Egyptian belief that the flood-waters of the Nile exude from H'py, the large-breasted Nile-god. Egyptian t establishes Nostratic /ho; and IE *nedo-, 'reed/cane', a plant which is bled for its sap, suggests Nostratic *t?. Arabic naDDa, 'ooze forth', narrows the final vowel to Nostratic *o. Uralic natta, 'mucus, slime', decides us in favor of Nostratic *a as the first vowel.

There is also IE *ned-, 'river', which apparently originated in a spring-fed stream.

In another publication, Bomhard bases his PAA "*nat'-/*n6t'-" on the Proto-Semitic "*nat'-ap^[h]- in Arabic naTafa, 'dribble, trickle, drip'; and also cites Egyptian ndf, 'sprinkle, moisten'. He, apparently, is not aware of 1) Pyramidic dfdf, 'drip', and dfdf.t, 'drop', which strongly suggest that the simplex is df; and 2) the prefix n- of Semitic (including Egyptian) roots which conveys a "passive and reflexive meaning" according to Moscati (1969:126); ndf should, therefore, mean something like 'to bled dry by dripping'. I can find no reference to ndf meaning 'sprinkle, moisten' but ndfdf and ndf mean 'cry tears'; and if we understand 'cry one's self out', it very well fits 'be bled dry by dripping'. And Arabic TafTafa, 'burn out (of a wick)', suggests something like 'ooze forth to exhaustion'. Incidentally, Arabic naTafa is probably related to Egyptian tf, 'saliva, spit'. As a result of these facts, we can only deplore Bomhard's methods of reconstruction; and substitute our own.

In IE, there appear to be another set of closely semantically related roots built on PL *NA-*?A, 'inside'-stative='be inside', which yields IE *sna:- when prefixed by s-mobile.

We therefore reconstruct the root as consisting of PL *NA-*T?O, 'inside-lump'='interior semi-liquid material'='mucus/sap'.

Accordingly, we amend:

DISCUSSION: We have discussed the root we believe must be reconstructed for 'time' above. Bomhard has it as:

The sign with which nw is written ('water-jar') is acknowledged by Egyptologists to also read jn.

We therefore reconstruct the root as consisting of PL *HHE-*NO-(*FA-), 'move-up-from-jar'='clepsydra(e) [with a float time-indicator]'='time(s)'.

No Altaic, Dravidian, nor Uralic cognate can presently be identified.

Accordingly, we amend:

(561)PN *hhen^w(/ow-) "**clepsydra(e), time" > PIE (2.) *en- "year"; *nu: (for **Henu), also shortened to *nu "now"; PAA *hhan(aw)- "**clepsydra(e), time" in Egyptian nw (for **jn(w) "time, hour"; S en_2,3 (for **in_x) "time, when"; in me-na-am₃ "when?" (for **mi₃-in_x-am₃ = "say time it is") (source: PL *HHE-NO[-FA-])

DISCUSSION: We have discussed the root we believe must be reconstructed for

We therefore reconstruct the root as consisting of PL *NHA-*FHA-(*KHO-)-[*¿A-/-*RO-], 'move-back-and-forth'+move-repeatedly='vibrate'='shimmer'(+'be covered' = 'glitter')[+'all'="shine"/+'very'='bright'].

The Sumerian sign which means 'wash' reads lah₃ and luh; the Sumerian signs lah₍₂₎ mean 'shine, sparkle; light'. I presume that there has been a confusion between the two, and that lah₍₂₎ should also have been able to be read **lûh_x.

The form we would normally expect Nostratic *laukh- to be **lûg in Sumerian but the form we actually find is **lûh if we are correct above. Readers may recall that in root #514, an expected Sumerian form **makâ appeared as mah. This was the result of PL *¿A, 'all', being appended to the root. It appears that this may also be the case here, where **lûgâ becomes lûh;and this would correspond with IE *luki-, 'light'. The Proto-Language word for 'wash' has an entirely different basis though the IE and Sumerian forms are similar. We will investigate it in the next discussion.

The Egyptian cognate for this root is nTr (for **nwTr), 'god'. This terminology is due to the stellar connections (or nature) of Egyptian divinities. The sign with with it is written was described by Gardiner as 'cloth wound on a pole, emblem of divinity' however the sign, when combined with the sign for 'bag (of linen)' means 'natron'. Natron is hydrous sodium carbonate, often found crystallized, i.e. it is highly 'reflective'. Before the discovery of metals, crystals would have been the most highly reflective objects the speakers of the Proto-Language might have terrestrially encountered. On this basis, and for other reasons (mainly, its plain form), I have regarded the sign as 1) either a metal flag or standard (pennant); or likelier in my opinion, 2) a halberd or pole-axe. Readers may judge for themselves if 'cloth wound on a pole' seems a more reasonable interpretation of the form of this sign.

No Altaic, Dravidian, nor Uralic cognate can presently be identified.

Accordingly, we amend:

(580)PN *la(:)w(/a(:)k^w-/ko¿-/kor^w-) "to vibrate, to shimmer, to glitter"/"to shine"/"to be bright" > PIE *leukh(y-)/*loukh(y-)-/*lukh(y-)- "to shine, to be bright"; PAA *law- "to shine, to gleam, to glitter"; in Arabic liwâ?-un "flag, standard"; in Arabic lâha (l-w-h) "to gleam"; in Arabic lâHa (l-w-H) "to shine (star)"; in Arabic lauH-un "metal plate"; Egyptian nT(-)r (for **nwT-r; cf. IE *leuk-ro-) "god (if = "very bright" {stellar divinity})"; S in lah₍₂₎ (for **lûh_x[?]) "to sparkle, to shine; light" (source: PL *NHA-FHA[-KHO]{-¿A-}/{-RO-})

DISCUSSION: The root for 'wash' upon which we touched briefly above is found in Bomhard's

which then murks the waters with:

To begin with, I can find no words with Arabic l-H- that refer to 'flowing, pouring, or moistening'. It also does not seem to be in Orel-Stobova nor in Ehret. Nostratic *lahh-, if it existed, should appear as PIE *la:- but there is no trace of *la:-, '**to pour out (liquids)' that I can find. This entry is, therefore, apparently a product of Bomhard's imagination based on his misreading of the significance of Sumerian lah₃ "to wash, to clean", which is probably simply the result of a confusion with **lûh. However, there is a Sumerian lah, which I believe is a legitimate reading: lah_(4,5), 'drive off; plunder, take away; fling (away)', which, in my opinion, corresponds to IE *(s)la/a:g^w-, 'take, grip, grab'; which also exists in Egyptian as nš (for **njš), 'expel, drive apart, put away'. Nostratic *x(a) results in IE *g^w, Sumerian h, and Egyptian š. We will revise Bomhard's #582 below.

The Sumerian sign which means 'wash' reads lah₃ and luh; the Sumerian signs lah₍₂₎ mean 'shine, sparkle; light'. I presume that there has been a confusion between the two, and that lah₍₂₎ should also have been able to be read **lûh_x.

The form we would normally expect Nostratic *louk?x- to be **lûk in Sumerian but the form we actually find is lûh. Readers may recall that in root #514, an expected Sumerian form **makâ appeared as mah; and in #580, an expected Sumerian form **lûk appeared as **lûh. This was the result of PL *¿A, 'all', being appended to the roots. It appears that Nostratic *hh(a) has been appended to the root here. This results in an IE form of *leugha: seen in OHG louga, 'lye'; here, a pre-Sumerian **lûk-ha has become lûh.

The Egyptian cognate for this root is found in rx.ty, 'washerman'. Instead of the correct root *rx (for **rwx as we shall shortly see), a faux root of rxt, 'wash', has been erroneously backformed. This may have been assisted by a form like **rwxj, an apparent weak root, for which an infinitive in -t would have been regular. That rx(tw is indicated strongly by the spelling of rx.ty, 'washerman', which is written with two 'white-fronted geese' combined into an overlapping monogram. One of the readings for the 'goose'-sign is r(j), a kind of 'goose'. The normal Egyptian dual is -w(j), so two 'geese' would read r(j).w(j). We believe, therefore, that this spelling indicates rx- should be understood as **rwx- or **rxwj, which would correspond with IE *low6-.

To summarize: the simplest Nostratic root for 'wash' (actually, 'rinse') appears to be *lo(:)w-. To this, the formant *-hh could be added: *lo(:)wa (:)hh-, 'wash-water'. Another expansion was represented by Nostratic *-k?x^w, 'hole', producing *lo(:)wa (:) k?x^w-, 'wash-hole'. And finally, to this, *-hh could also be added: *lo(:)wa (:)k?xohh-, 'wash-hole-water' = 'lye' or some other cleaning agent.

We therefore reconstruct the root as consisting of PL *RHO-*FHA-(*K?XO-)-[*HHA-], 'rise'+move-repeatedly='bob, move up and down'= 'rinse'+'hole'='washing-place at river'+'water'='washing agent'.

Nostratic *l has three sources in the Proto-Language: NHA, which will be Nostratic *l; N(H)E, which will be Nostratic *l²(:); and RHO, which we will indicate as *l³.

No Altaic, Dravidian, nor Uralic cognate can presently be identified.

Accordingly, we amend:

(581)PN *l³o(:)w-(/wa(:)hh-/wa(:)k?x^w-/wa(:)k?xohh-) "to wash" ("wash water"/"washing hole"/"washing agent") > PIE *lo(:)u/w6-/wa(:)gh/ugha: "to wash" ("wash water"/"washing hole"/"washing agent"); PAA *lawag- "to wash"; in Egyptian **rwx "to wash"; rxt (for **rwxj.t) "to wash"; rx.ty (for **rwxj.tj "washerman"; S in luh "to wash" (source: PL *RHO-FHA[-HHA/-K?XO/K?XO-HHA])

No Altaic, Dravidian, nor Uralic cognate can presently be identified.

Based on a reconstruction of PL *NHA-*HHE[?]-(*XA-)-, 'move-back-and-forth'+move-up-from(?)+'done-energetically'='shake-off-vigorously(-to-clean'),

(it should be noted that, in the absence of a Semitic cognate, the medial 'laryngeal' can only be a 'guesstimate' based on the semantics of the root)

we amend:

(582)PN *la(:)hhex-)[?] "to shake off vigorously (to clean), to clean off by shaking" > PIE *(s)la:g^w- "to take, to grip, to grab"; PAA *lahhax- "to rid of "; in Egyptian **nš (for **njš) "to expel, to drive apart, to put away ("to shake off")"; S in lah₃ "to clean (for "to shake vigorously")"; lah_(4,5) "drive off; to plunder, to take away; to fling (away) ("to shake vigorously")" (source: PL *NHA-HHE-XA)

DISCUSSION: In my opinion, Bomhard has reconstructed an incorrect root-vowel based on his incorrect equivalencies. Although the IE form, *lew- could arise from any Nostratic vowel, the variation of lu/u:- suggests to me that we are dealing with an original **le:w- which has been shortened. This would indicate Nostratic *e for:

We therefore reconstruct the root as consisting of PL *NHE-*FHA-(*THO-)-, 'slippery-move-repeatedly'='grimey/dirty'(+'large animate definite plural' = 'dirty creatures').

If my reconstruction of the Nostratic parent-form is correct, we would expect a Sumerian reflex of **liu, which developed into **lü.

No Altaic, Dravidian, nor Uralic cognate can presently be identified.

Accordingly, we amend:

(579)PN *l²e(:)w-(/wa(:)th^w-) "to dirty ("dirt") > PIE *le:u(t)- "to make dirty ("[infestation of creatures living in ]dirt")"; PAA lawat^w in Arabic lâTa (l-w-T) "to stain, to tarnish, to soil, to sully", lauTâ "stain, blot, spot"; S lu₃ (for **lü₃ from **liu) "to become dark (= "to become dirty")" (source: PL *NHE-FHA[-THO])

DISCUSSION: In reconstructing the Proto-Language, it became apparent to me that there were a number of roots with IE *l, for which Egyptian had n. This suggested to me that, in at least one branch from Nostratic, an original Nostratic *n(:) [from PL *N(H)] had been retained but selectively denasalized in other branches since some IE *n also corresponded to Egyptian n. For IE, apparently all aspirated PL NH became *l but *n was retained for PL NA and NO; PL NE also became IE *l. The cognates in the entries discussed below indicate that for Altaic, Dravidian, and Uralic, PL *NHE and *NHO remained *n(y)/*ñ. It must be emphasized that these palatalizations are not due to a following *e or *y but are an attempt to preserve the semantic integrity of words from the Proto-Language beginning with an aspirated *NH. This data gives further support for the reconstruction of the series of PL aspirated nasals (*MH, *NH, *QH).

Understanding this allows us to include the proper PAA cognate, which Orel-Stobova have as "1650 *lam- "be soft"", since all aspirated PL *NH becomes l in Semitic. Bomhard must pass it by.

There are two roots, the reflexes from which Bomhard has mixed together:

and

If one looks at the IE root Bomhard proposes, it is obvious that 'pliability' is not the idea primarily involved but rather 'bowing' to show respect. And for the PA cognate, Starostin has *n^yuma-, 'warm, soft, mild', indicating Nostratic *o rather than *e.

We therefore reconstruct the first root as consisting of PL *NHO-*MHO-, 'slack'+ 'always-do'='soft'.

In regard to the second entry, a Sumerian sign meaning 'break apart' has the reading lim(m)u-; and we provisionally assume that this meaning may be associated with this reading.

We therefore reconstruct the second root as consisting of PL *NHE-*MHO-, 'come-apart'+ 'always-do'='stretchable'.

Accordingly, we amend:

(576)PN *lo(:)m^w- "to be slack; soft" > PIE (1.) *lem- (for **lo:m-, later **lom-) "(to break apart; broken apart,) soft"; PAA *lam- "to be soft"; perhaps in Egyptian nm(-)' "to be one-sided, partial (if "soft of arm"[?])"; PU *n^y8m3^-1- "soft" (Rédei); PA *n^yuma- "warm, soft, mild"; S lum "to soften" (source: PL *NHO-MHO-)

and amend again:

(577)PN *l²e(:)m^w- "to break/come apart easily; very fragile" > PIE (1.) *le(:)em- "to break apart; broken apart(, soft)"; (2.) *le(:)em- "to gape"; PAA *lam- "to gape"[?]; perhaps in Arabic lamma "to be seized with madness (if "to gape"[?])"; probably in Egyptian nm "to go wrong (plans); to rob, to steal"; perhaps in Egyptian nm.t "slaughter-house"; PU *n^yäm3- "soft" (Rédei); PD *ñim- "to stretch"; S **lim(m)u- "to break apart[?]" (source: PL *NHE-MHO-)

DISCUSSION: At first sight, it might look like Bomhard's #574 is similar in origin to the two roots discussed immediately above but, a signpost for a mistaken reconstruction, is that Bomhard is able to provide no PIE, PAA, PA, nor Sumerian cognates.

An IE root that matches the meaning is (1.) *kem-, 'press together', which we connect with Egyptian qm(-)3, and Sumerian kam_(1,2), 'grasp, grip'. From Egyptian qm(-)3, we know that the first consonant must be Nostratic *nkh (PL QH); and the PFU and PD forms tell us the first syllable must be *nkha- (PL *QHA-, 'hump-up'). This means that the Sumerian cognate must be amended to (n)k²am_(1,2); and the IE form represents *(n)kem-. Obviously, the root means 'bend something over'.

It is, therefore, apparent that though Afroasiatic, Egyptian, IE, and Sumerian continued the *k-element of *QH, Uralic and Dravidian maintained the *n-element, and palatalized it, as above, to maintain semantic integrity.

We therefore reconstruct the first root as consisting of PL *QHA-*MO-, 'hump-up'+ 'slap/pound'='bend over so as to be together'.

I can find no suitable Altaic cognate at present.

Accordingly, we amend:

(574)PN *nka(:)m^w- "to bend over so as to be together, to press together" > PIE (1.) *kem- (for **(n)ka(:)m-, later **kam-) "to press together (by doubling over)"; PAA *qam- "to press together" (Ehret has the root as #447 **k^w'am- "to curve") in Arabic qamaza "to pick up with the fingers", qamasha "to pick up refuse", qamaTa "to swaddle", qama¿a "to empty a skin"; in Egyptian qm(-)3 "to hammer out"; PFU *n^y8m3^-2- (for **n^yam3^-2-; see Rédei, who discusses the irregular vowel correspondence) "to press (together), to squeeze"; S kam_(1,2) (n.) "grasp, grip"; PD *ñam- "to press, to squeeze, to crush, to pinch" (source: PL *QHA-MO-)

DISCUSSION: The correspondent Nostratic *la(:)w- to *le(:)w- and *lo(:)w- discussed above is Bomhard's

If one looks at the IE root Bomhard proposes, it is hard to see how he was able to interpret it as "to bend, to twist, to turn" when the meaning quite clearly appears be centered around the idea of 'loosening' by whatever method of manipulation.

With extensions in *-s-, the IE root mentioned above means 'lie' and 'lying, deceptive'; and, without the *-s-, Sumerian lu₅ means 'liar; lie, deceive, false, treacherous'. I presume the idea is being 'loose with the truth', and consider the Sumerian word a cognate.

If our methodology was correct above, we should expect a Dravidian cognate in the form of **ñu and Altaic/Uralic cognates in the form of **n^yu — if they exist.

No suitable Altaic nor Dravidian cognate seems to be presently ascertainable.

We therefore reconstruct the root as consisting of PL *NHA-*FHA-, 'move-back-and-forth'+ 'move-repeatedly'='loosen'.

Accordingly, we amend:

(584)PN *la(:)w- "to loosen; loose" > PIE (2.) *lew- (for **la:w-, later **law-) "to loosen"; PAA *law- "to loosen, **to lie/deceive"; PU possibly in *n^yu(-)j3- "to flay, to pull off the skin or pelt"[?] (Rédei); S lu₅ (n.) "liar"; (v.) "to lie, to deceive"; (adj.) "false, treacherous" (source: PL *NHA-FHA-)

DISCUSSION: The other major negative of Nostratic is listed in Bomhard's

If one looks at this entry, one is stricken by the fact that Bomhard has proposed a reconstruction that is totally meaningless since it includes every vowel possible — what can explain anything, explains nothing. In addition, setting every vowel side by side highlights a peculiarity of Bomhard's proposal; *i and *u are front and back closed vowels, accompanied by allophonic open vowels: *e and *o. Strangely, the central closed vowel, *6, takes the second position when, to be consistent, it should take the first position: i.e. *6/*a.

Now, in Bomhard's #564, he lists "PN *na-/*n6- 1st person personal pronoun stem", and in #570 "PN *na-/*n6-, *ni-/*ne-, *nu-/*no- demonstrative stem". This kind of ambiguity cannot be tolerated by most languages. And, we would reconstruct Nostratic *na, 'one', for both #564 and #570. On the basis of avoiding unnecessary ambiguity, we will rule out *na/*n6 as the "negative/prohibitive particle". We will shortly be illustrating that Nostratic **ne/**ni may not exist because all PL NE became *le in Nostratic. That leaves only Nostratic (*nu/)*no as a possibility, and the PL meaning of this word (*NO), 'put inside, discontinue', explains its negative connotation.

This word could be combined with Nostratic *wa, 'again', to produce 'never'; and this is probably the form (*now) we see in "PK *nu prohibitive particle" and Egyptian nw — if it exists. It is not recognized in Old or Middle Egyptian as a negative; and the negative has survived into Coptic only in a stress-unaccented form (n). However, the word is most frequently written with a sign depicting 'two connected arms with palms turned down in a gesture of negation or rejection'. There is an Egyptian word nj with 'one arm with palm turned downward' as a determinative that means 'reject, refuse'; as the Egyptian favored graphic puns, 'two arms with palms turned downward' might well have been a way if indicating **nj.w (-w[j] is the normal suffix of the dual); in which case, it would correspond to IE *nejew, 'never'. Presumable, this would represent a adjective-form of the negative (*NO-*¿E-, 'put-inside'+ '-like'='absent'), seen also in IE *ney. Bomhard's Egyptian n3 is also not attested early, and probably represents a late spelling to indicate a lengthened vowel.

Bomhard mentions "PU *ne negative particle". I am unable to find anyone who recognizes this for Uralic, for which the normal negative is usually reconstructed to be *e, which I believe to be a result of *HHE-*¿E-, 'leave'+'-like'='leaving/going-away/(soon to be) not-present'. The simplex (*HHE) of this word can be seen again in the Egyptian negative jw.t(y), 'who/which is not'; in Egyptian jw, 'be boatless', and in IE (1.) *eu-, 'lack, empty' (*HHE-*FHA-, 'leave'+'move-repeatedly'='cause-to-be-lacking' (Bomhard has this root (almost) correctly as #409 "PN *hhiw-/*hhew- "to lack, to stand in need, to be in want" = Nostratic *hhew-). We will not include *ne.

This leaves the vocalism of Sumerian na, 'not, prohibitive prefix', to be explained. The sign for na also reads nu₈; and I believe it probably it probably represents Nostratic *now, 'never' (nû₈), particularly in view of the fact that it has an affirmative use as well. The affirmative use could represent an abbreviated Nostratic *hhenow, 'now', mentioned above in #561, similar to IE *nu/u:.

'Never' seems a good semantic choice for injunctions like: na-an-mu-mu₂-un, 'do not start a quarrel'. It makes much more sense to understand this as a universal prohibition ('never start quarrels') than as an instruction for a particular occasion. Similarly: nam-mi-gul-e, 'no one shall destroy it' — or, 'no one shall ever destroy it'. On the other hand, na-affirmative seems to just set the stage in the same way we use 'now' in English: nam-ta-e₃, '(now) he went out'; and nam-mi-gub, '(now) he set (his foot on the ship)'.

Sumerologists cannot be faulted for reconstructing na for these usages since both, when it is combined with the verbal prefix mu-, it is written nam; and nam has no recorded variant reading of **num (though nam₄ also reads num). If we were to assume that the bird pictured was an 'ostrich', it might be possible to think of it as being called by the Sumerian equivalent to njw, 'ostrich', which, of course, would have the same form Nostratic *noyew, 'never' (for 'ostrich', PL *NO-*¿E-*FA-, 'basket'+ '-like'='vessel'+formant of bird names='ostrich'. Before pottery; baskets, bird-eggs, and skulls served as containers. This would have produced Nostratic *noyew- also.), and what we have supposed might be the correct reading for the Egyptian sign with 'two arms' (**nj.w). There is also Egyptian njw, 'bowl' — quite probably: 'ostrich[-egg]'.)

Though we may never be able to offer conclusive proof, I believe it is quite possible that the Sumerian sign reading nam may have also read nü_x or num_x.

Similarly difficult to elucidate is the combination of the prohibitive with an n-prefix to na-an- as in e.g. na-an-mu-mu₂-un above. In such examples, we simply do not know the meaning of the prefix nor its original form so speculation is just that: speculation.

Perhaps of greater interest is the fact that the sign for the Sumerian negative we have identified, nu, also reads la₃, recalling the normal Arabic negative lâ and ul2, recalling the Akkadian negative ul and Hebrew ?al. Could nu have been read la₃ or **al_x in certain contexts as a Sumerian negative? Or are these loanwords from a Semitic language?

I propose for this "negative" a reconstruction of *?A-*NHA-, 'forehead'+'move-back-and-forth'='to shake the head, to indicate unwillingness or inability, be incapable'.

Bomhard has this root in his #449:

Orel-Stobova have this root as "149 *?Vl- "to be exhausted"; and this is very probably the basal semantic idea, backed by Arabic ?alâ (?-l-w), 'be unable, incapable' (PL *?A-NHA-FHA, 'shake the head repeatedly').

I believe this 'potential negative' can also be found in Egyptian as nn. It is understood that a tripled writing of the 'wave'-sign (n) reads mw, 'water'. But what is not presently understood is that some doubled writings of n do not read n-n but rather **jn. For example, the Egyptian word for 'be weary', which is normally transcribed as nn(y) is fairly frequently written with n-n, usually preceded by doubled 'rush with shoots'. This sign ('rush with shoots'), I regard as the Egyptian representation of Nostratic *hhal^w- (PL HHA-NHA 'push-across'+ingressive='start to go towards, change', become' (Bomhard's #380) — which would also result in Egyptian jn [IE (2.) *al-, 'grow']. It can, at the same time, hardly be seriously doubted that nn (i.e. **jn) represents Nostratic *?al-, 'be tired, incapable', seen also in IE (2.) *el- (for *a:l-, later *al-), 'rest; be tired'.

And, as a final indication of this proposal, it is known that Egyptian nn is the preferred negative for present or future reference where potentiality or capability is of interest.

No suitable Altaic nor Dravidian cognate seems to be presently ascertainable.

We therefore reconstruct the simplex root as consisting of PL *NO-, 'put-inside, discontinue'='stop'.

Accordingly, we amend:

(562)PN *no(w/y) "to put inside, to discontinue, to stop" negative ("never"/"absent"="not") > PIE (2.) *ne(y) negative ("not"); PK *nu prohibitive particle; PAA *na negative in Egyptian n(j/y) "not"; PA in *a:(-)ni- "not"; negative verb; S nu negative; na (for **nû₈) prohibitive ("never[?]") (source: PL *NO[FA/¿E])

We can complete the discussion of the other roots discussed above. As for #449, I must say that I am unaware, and know of no source which identifies Sumerian li as a negative; accordingly, I will omit it. However, IE (2.) *el- does show what appears to be an adjectival form in -*y (*eley-). This could account for Hittite li-e if we assume it has been shortened from **elê as Arabic lâ appears to have been shortened from?alâ. Also, I am willing to include Bomhard's PA *üli- (in Starostin's form: ulE), which may be a combination of the negative in root #449 preceded by the negative in root #409.

And also, based on the foregoing discussion, we can amend:

And Bomhard's #409, mentioned above, can also be amended to:

and also amend:

DISCUSSION: Here Bomhard has conflated two important roots: the first, in which we will amend #409, means simply 'have intercourse, beget'; the second has to do with 'bending over', which can also be construed as the passive partnership of copulation as well as the travails of the birth process. What is rather distressing is that the IE root is reconstructed as *g^en-, with a palatalized initial *g. Bomhard's reconstruction with Nostratic *a(/*6) cannot hope to explain this palatalization. And he also seems unaware of Dravidian *ken.-, 'have sexual intercourse', which correlates with *g^en- perfectly."

There is a slight problem, however. IEists seem to be of the consensus that all attested IE forms can better be derived from IE *g^enH₁- than from *g^en-, an opinion which may or may not be correct. In deference to the consensus, I will reconstruct the IE root as derived from Nostratic *k?eno?-, 'begotten'.

Two sign reading DUMU and DIŠ, together read gina(n). When combined with GA, 'milk', the resulting combination means 'unweaned child'. I propose that the 'child' element is contained in DUMU-DIŠ, reads **gin-na, 'what has been begotten' = 'child'.

We therefore reconstruct PL *K?E-*NO-, 'erect-penis-put-in'='to have intercourse'; and, further:

PL *QA-*NA-, 'semi-circular'+ingressive='start to be bent over'.

I can find no suitable Altaic cognates at present for either root.

Accordingly, we amend:

(275)PN *k?en^w- "to have intercourse, to beget" > PIE in (1.) *g^enH₁-/*g^onH₁-/*g^NH₁- (*g^eno:- from *k?eno?-) "to beget (from "begotten")"; PAA *kan- "to beget" Arabic in kanaba "to stow a thing in a bag", kanara "to press a thing in a bag, to stick a spear"; in Egyptian kn(-)s "pubic region"; PD *ken.- "to have sexual intercourse"; S **gin-na (n.) "child" (source: PL *K?E-NO-)

and

(Addition AG)PN *nkan- "to be bent over, to submit, to be crouched giving birth" > PIE *gen-/*gon/*gN "to press together, to pinch, to crease, something pressed together, something balled up"; PAA *qan- "to be bent over" in Arabic qanata "to humble one's self"; in qaniya "to be hooked"; Egyptian qnj "to overpower"; in qn(-)b "to bend, to bow, to incline one's self, to subjugate"; PD *kan- "to bear or bring forth, to beget, to bear children; (n.) young of various animals, young child"; S (n)g[~]³an- "to bear, to bring forth, to give birth to" (source: PL *QA-NA-)

DISCUSSION: What is again rather distressing is that the IE root is reconstructed as *g^en-, with a palatalized initial *g. Bomhard's reconstruction with Nostratic *a(/*6) cannot hope to explain this palatalization either.

We therefore reconstruct PL *QA-*¿A-*NO-, 'semicircular'='cupped (hand)-eye-put-in'='look at intently'='recognize'='know how to'.

There are other compounds utilizing PL *¿A, 'eye': e.g. **T?SA-¿A, 'long-eye' = 'stare', in IE *dheyeH-, 'see, look' (Orel-Stobova have it as #622 *da¿-/*di¿- "look, know"); and **SA-¿A-(FA-), 'strong-eye'(+'do-repeatedly') = 'watch', in IE *se:w-, 'notice'; Egyptian zj(-)3, 'perceive'; possibly Sumerian sa₄ (for **sâ₄), 'name; call by name'; and possibly PA *se:ja(u)- (Starostin), 'consider, count'.

I can find no suitable Altaic, Uralic, nor Sumerian cognates at present for this root.

Accordingly, we amend:

(295)PN *nka¿an^w- "to recognize, to know (how)" > PIE (2.) *g^yen-/*g^yon/*g^yN (*g^eno:- from *nka¿ano?-, "known") "to recognize, to know (how to)"; PAA *qa¿an- "to observe closely"; Egyptian in qnj (for **qjn(-)j) "to be able"; PD *kan.- (for **ka:n.-) "eye"; *ka:n.- "to see, to observe, to consider"; (?)PE *kangiRci- "to understand" (source: PL *QA-¿A-NO-)

DISCUSSION: It is obvious from the table on page 51 of his book that Bomhard wishes to reconstruct affricates for, at least, the apical series of Nostratic consonants, for which he is to be commended. What is not so obvious is that, to designate (presumably) [dz{h}], he uses a symbol which in the IPA designates simple [zh], which we will transcribe here as *Z. Similarly, following the erroneous pattern he has established in the stops, he reconstructs *c^[h] (presumably [tsh]) and *c'. Why he has opted for shibilants rather than sibilants escapes me completely. I will proceed to show that there is no difference in the reflexes to Bomhard's PN *Z and *c', and hence will prefer my own reconstruction: t?s with sibilant, which corresponds to IE *dh; and Semitic *d before Nostratic *a and *e, *Z (dotted z) before Nostratic *o. In a previous publication, the roots here with *Z are spelled *d^z so I presume, in spite of the *Z, that something like simple [dz] is what is (was?) desired to be indicated. Bomhard's *c^[h] (formerly *t^s[h]) is my Nostratic *ths, which corresponds to IE *t(h); and Semitic *dh before Nostratic *a and *e, *S (here, dotted s) before Nostratic *o.

DISCUSSION: What is quite disappointing is that Bomhard, in another publication, cites no Arabic cognate, when one is easily found, and mentions Egyptian z(j)3, which is totally inappropriate since Egyptian /3/, in any early spelling, is a kind of [r*sa:?/?ay[ar/yar]-, corresponding to IE *sa:(ir(o)]-, 'pain, sickness, damage'; and Sumerian sa₅ (for sâ₅ ), 'red' (PL *SHA-?A-[/¿E][/-RHA], 'pinkish'+stative[+-like'][+color]='reddened').

The IE cognate, *dhe:-, should tell us that the Nostratic vowel in this root should be *e but in view of Altaic *Za:-, 'lie, be stretched out', it looks very much as if the IE vowel has been incorrectly reconstructed; accordingly, we amend to **dha:- which then relates to *dha:w-, 'strangle' (if 'stretch the neck').

The expected Dravidian cognate would have the form **Za:-; and Dravidian *Ca:-, 'stretch out', should rather correspond to *T?SA-*?A-, 'stand-up'+stative='be erect', that we see with s-mobile in IE *sta:-, 'stand'.

We therefore reconstruct PL *T?SA-*?A-, 'long'+stative='be long, be stretched out, be prone'.

I can find no suitable Uralic cognate at present for this root.

Accordingly, we amend:

(172)PN *t?sa?- "to be stretched out, to be prone" > PIE (3.) **dhe:- (for **dha:-) "to fade, to be exhausted (and, as a result, "to be lying down"); PK (*Z?-in-) *Z₁-in- "to lie down, to go to sleep"; PAA *da?- "to be still"; Arabic da?da?a "to still"; tada?da?a "to be still"; PA *Za:- "to lie, be stretched out"; S za₃ (for **zâ₃; another reading of zag) "border (if = "what stretches out")" (source: PL *T?SA-?A-)

DISCUSSION: A semantically somewhat closely related root is Bomhard's:

DISCUSSION: Bomhard makes two mistakes — rather consistently: 1) he defines the meanings of the Nostratic roots so broadly so that correct cognates cannot be identified; and 2) appears never to go to the primary sources for information. We have a typical example in his handling of this root of both problems.

Firstly, as we shall see from the semantic analysis of the root, it means 'release repeatedly', i.e. 'flow away'. To expand the definition to 'to pass away' opens the door to false identification with Arabic zâla (z-w-l), 'go away' (which is cognate with the root in IE *sew6-, 'let loose, release'), when the correct cognate is Arabic dâla (d-w-l), 'elapse (of time), flow'. Bomhard mistakenly lists Egyptian zw3, 'pass', with which there are two problems: 1) the word is s(-)w3 not zw3; and 2) it is a s-causative of w3, 'fall': 'cause to fall'. There is no Egyptian cognate for this root though an Egyptian cognate exists for the second root to be reconstructed: Dw(j), 'evil, bad' = 'ill' (Bomhard has mixed this up with a totally unrelated root in his root #154 "PN *t'^yaw-/*t'^y6w- "bad, evil""; this is PL *T?E-(-*FA-[-*SO-]), 'under'+'do-repeatedly'+'pull'='pull-under'='substandard, low-quality'), seen again in his #139 "PN *t'aw-/*t'6w- "to leave, to go away; to send forth, to let go, to chase away, to release"; IE (1.) *dew-, 'submerge'; in Arabic tâkha, 'dip into a soft substance'; tâ¿a, 'dip bread in [butter]').

Secondly, another similar root must be distinguished.

We therefore reconstruct PL *T?SA-*FHA-, 'long'+'move-repeatedly'='last, pass'.

and, for the second root:

We therefore reconstruct PL *T?SA-*FA-(*?A-[*¿E-]), 'long'+'do-repeatedly'='lie down often, be sick'(+stative='be dead'[+-'like'='be dying').

Bomhard actually has this root though he has misconstructed it as:

I can find no suitable Altaic, Dravidian, Sumerian nor Uralic cognates at present for these roots.

The first root, #171, in the form Bomhard cites it (Pokorny's *dhwi: from (2.) *dheu-), appears to be merely an extension of meaning for Pokorny's (1.) *dheu-, 'run out (of liquids)', which Bomhard has as a separate root:

I am incredulous at this entry. Bomhard seems to equate 'run (of liquids)' with 'run (on legs)' and with 'run (between the legs)'! Bomhard's ad hoc correspondences call for Arabic D = Nostratic *d^y, the confusion of palatality with fricativity. And the Semitic "cognates", cited in another publication, are in the semantic range of 'melt' and 'wilt'. Obviously, any 'flow' will do. Bomhard's cited Arabic dhawâ, 'wither, wilt, fade', should be referred to IE *ta:u- (for *thau-), 'rot, disintegrate', rather to any IE root with initial *dh. Frankly, I will be kind when I say that this entry is ludicrous.

We therefore reconstruct PL *T?SE-*FA-(*¿E-), 'release'+'do-repeatedly'='run out'+progressive='be running out, emptying'

Accordingly, we amend:

(145 & 171)PN *t?sew-(/way-) "to run out (to be emptying)" > PIE (1.) **dheu-/**dhou-/**dhu- "to run"; (2.) **dheu-/**dhou-/**dhu- "to dwindle"; dhwi:- (from **dhewey-) "to be unconscious"; PK in *Z₁-w-el- "old"; in *Z₁-w-en- "to grow old"; PAA *daw- "to pass"; Arabic in ?indâla (d-w-l) "to emigrate (people)" (source: PL *T?SE-FA-(-¿E)

and

(Addition AH & 85)PN *t?saw-(a?-/a?ay) "to lie down often, be sick(, to die/to be dying)" > PIE (2.) *dheu-/*dhou-/*dhu-; *dhwe:i- (for **dhwa:i-) "to fade, to be exhausted, to become unconscious, to die"; PAA *daw- "to be sick"; Arabic in dâ?a (d-w-?) "to be ill, diseased"; in dawiya "to be diseased"; Egyptian Dw(j) "evil, bad"; PD *tav- "to die" (Starostin); *Ca:v-- (for **Zav-[?]) "to be ill, to die"; S zu₄ (for zû₄) **"dead (the sign which means "dead" [UŠ₂] also reads zu₄)" (source: PL *T?SA-FA-[?A-/?A-¿E])

DISCUSSION: In another publication, Bomhard lists Egyptian zrm, 'flood, torrent'; the Erman-Grapow Wörterbuch knows only srm.t and srmm, 'a kind of body of water'; and both are not attested until Late Egyptian times. Of course, there is s(-)rm, 'cause to weep'. A zrm.t, 'kind of drink, is recorded; and with the usual later confusion between s and z, a spelling of zrm.t for srm.t is recorded. When we look in Budge's dictionary, we do find zrm, 'flood, torrent', but what modern researcher would prefer Budge's interpretation to those of Erman-Grapow?

A second problem is that, if Sumerian zar is a correct cognate, Egyptian s/zrm cannot be because all Egyptian r shows up in Sumerian as l. What is most likely is that Late Egyptian s/zrm is a simple loanword from Hebrew zerem, 'flood of rain, downpour'. A third problem is that, as we have seen above, Egyptian D not s or z corresponds to Sumerian z, which probably was phonetically [ts], reflecting the affricate that *Z ultimately represents ([*dz] or [*ts]). I opt to keep Sumerian zar, and must therefore reject **zrm and Hebrew z-r-m.

Actually, the Arabic (hence, Semitic) cognate is easy to find once one employs the correct correspondences: darra, 'flow abundantly', and dara?a, ' rush (torrent)'. Hebrew zerem is hardly necessary when Hebrew d-r-r exists in the meaning 'flow rapidly, forcefully'.

Hebrew zerem is instead cognate with IE *ser-, 'stream, move rapidly and powerfully", corresponding to Sumerian sar, 'run' (PL *SA-*RHE-, 'strong-rain'='stream'), found in Arabic zarî¿-un, 'field watered by rain' (Bomhard attempts to analyze this Nostratic root in #163 "s^yur-/s^yur-", which we will discuss and amend after this discussion).

And the alveolar r of the Dravidian cognate, which I accept, tells us that the final vowel of the root was Nostratic *e while its long a: probably indicates the former presence of *? — as in one of the Arabic cognates — for a final.

We therefore reconstruct PL *T?SA-*RHE-, 'long-rain'='(flash-)flood'.

I can find no suitable Altaic nor Uralic cognates at present for these roots.

Accordingly, we amend:

(176)PN *t?sar^y-(/re:?-/re:wa:-) "to (flash-)flood(to be flooded/to spill out") > PIE (4.) **dher-/**dhor-/**dhR- (**dhereH-/**dhoreH-/**dhRH-/**dherew-/**dhorew-/**dhRu-) "to jump, to cover (sexually), to stream rapidly (to be flooded/to spill over)"; PAA *dar- "to rush"; Arabic in darra "to flow abundantly"; in dara?a "to rush (torrent)"; PD *ca:r- (from *Zari:-[?]) "to flow, to run (off or out), to issue, to drop or ooze out, to drizzle"; *Zo:r- (from *Zariv-[?]) "to leak"; could this shift from *Z to *c possibly be due to the effects of a Dravidian "s-mobile?; S zar "to run, flow, leak, or spill out; to spring forth, to issue (from), to flow or gush forth; to bubble over"; PE *caRvar "current" (cf. Central Alaskan Yupik caRvaq "current, rapidly flowing, stream", caRv6-, caRvaR- "flow [of current]") (source: PL *T?SA-RHE-[*?A-, stative='be-flooded'/*FHA-, 'move-repeatedly'='be-flooded'])

DISCUSSION: In the root discussed below, Bomhard demonstrates a misunderstanding of a fundamental nature. At the Nostratic level, there can be no phonemic *s^y as an initial because the vowels which would produce glides in the daughter languages are still present. Therefore, Bomhard's series of palatalized and velarized consonants is just a misleading extravagance that adds nothing to the correspondences, which are regular without the modification of glides.

In order to justify PD *co:r- and Sumerian šur (for **šür), Bomhard has invented a Nostratic phoneme that never could have existed.

Sumerian šur corresponds to IE *swer-, 'fester, suppurate' (*SE-*FA-*RHE-, 'bodily-excretion'-plural-'fall'='suppurate'), Nostratic *sewar^y.

Dravidian *co:r- corresponds to IE (8.) *(s)ter- [for **(s)ther-], 'excrete' ([*SA-]*T?SO-*RHE-, [strong-]'arm'='fecal-roll'-'fall'='excrete'), Nostratic *(sa)t?sor^y), Bomhard's root #76, "PN.*dur-/dor- (which I would amend to amended *t?sor^y-).

In regard to Dravidian, we will see in the next entry to be discussed (#170) that Dravidian has a natural response to Nostratic *s(h), namely *s.

A slight complication is that, again, two Nostratic roots with similar meanings seem to be involved in this entry.

We therefore reconstruct PL *T?SA-*RHE-, 'long-rain'='(flash-)flood'.

I can find no suitable Altaic nor Uralic cognates at present for this root.

Accordingly, we amend:

(163)PN *sa:r- "to overflow" > PIE *ser-/*sor-/*sR- "to move quickly, to run, to flow"; *ser-p^[h]-/*sor-p^[h]-/*sR-p^[h]- "to creep, to crawl"; *sr-ew-/*sr-ow-/*sr-u- "to flow"; PAA *sar- "to surge, gush, flow, spring, or spread forth"; Arabic in sariba "to overflow"; Egyptian in z3b "to drip"; PD *sa:r- to be left over, in excess; S perhaps in šar₂ (for **sâr₂, another reading of the same sign) "to beg, to implore"[?] (source: PL *SHA-RA-)

But we must also reconstruct PL *SA-*RHE-, 'strong-fall'='stream'.

I can find no suitable Altaic, Dravidian, nor Uralic cognates at present for this root.

(Addition AI)PN *sar^y- "to stream" > PIE *ser-/*sor-/*sR- "to move rapidly and powerfully"; *ser-p^[h]-/*sor-p^[h]-/*sR-p^[h]- "to creep, to crawl"; *sr-ew-/*sr-ow-/*sr-u- "to flow"; PAA *sar- "to surge, gush, flow, spring, or spread forth"; Arabic in zarî¿-un, 'field watered by rain'; Egyptian in z3 "to betake one's self to"; S sar "to run" (source: PL *SA-RHE-)

DISCUSSION: Two Nostratic roots with similar meanings seem to be involved in this entry.

We therefore reconstruct PL *SA-*¿E-*RE-, 'sinew-like'='cord'+'apply'='tie'.

Based on the correspondences I have established over many roots, Sumerian š can result from Nostratic *s(h)e (probably [sh]) or *x(h)e (probably [ç]), which were conflated. A Sumerian e resulting from the contraction of *V+*y should not affect the same change (from *s to *š).

I can find no suitable Altaic nor Uralic cognates at present for this root.

Accordingly, we amend:

(170)PN *sayer^y- "to apply a cord, to tie, to fasten" > PIE (4.) *ser-/*sor-/*sR- (for **syer-; cf Greek hen-eíro:, "fasten to") "to line up together, to fasten"; PAA *zayar- "to tie"; Arabic zarra "to button a garment"; zayira "to put someone in a strait"; Egyptian z3 "to protect"; PA *sir- "sinew, tendon"; S šer (for ser₄, another reading of the same sign) "to tie, to bind", šer₂(-šer₂) (for ser₂, another reading of the same sign) "to tie, to bind" (source: PL *SA-¿E-RE)

But we must also reconstruct PL *SHA-*¿E-*RE-, 'immobile-like-apply'='chain up in a line'.

I can find no suitable Altaic nor Uralic cognates at present for this root.

(Addition AJ)PN *sa:yer^y- "to chain up in a line" > PIE (4.) *ser-/*sor-/*sR- "to line up together, to fasten"; PAA *sayar- "to line up"; Arabic sâra (s-y-r) "to travel (to form a column=chain)"; sâra (s-y-r) "to establish a custom (to institute a recurring, chain act)"; saiyara (s-y-r) "to relate ancient stories (chain-recital)"; in ?asara "to tie (to chain together)"; Egyptian z3 "cattle-hobble"; S šer₃-šer₃ (for sêr₃sêr₃, another reading of the same sign) "chain", šir₃-šir₃ (for sêr₃-sêr₃, another reading of the same sign) "band, chain"; PD **se:r- "to unite, to join" (source: PL *SHA-¿E-RE-)

DISCUSSION: We mentioned above that Egyptian Dw and its derivatives belong to another root altogether [(Addition AH & 85)PN *t?saw-(a?-/a?ay) "to lie down often, be sick(, to die/to be dying)"]. In my opinion, however, the root contained in IE *dus- (Pokorny's entry for Bomhard's "*t'ews-/*t'ows-/*t'us-") has a different origin; and Bomhard has totally misconstrued its basic meaning (v. amended #154 below).

We therefore reconstruct PL *T?E-*FA-(*SO-), 'under'+do-repeatedly'[+'pull']=['pull]-to be put under'='substandard, low-quality'.

I can find no suitable Altaic nor Uralic cognates at present for this root.

Accordingly, we amend:

(154)PN *t?ew(was^w)- "to be submerged, to be at a low level(to submerge)" > PIE (1.) *dew-/*dew-/*du- "to submerge"; several additional cognates are listed mistakenly by Pokorny under (3.) *dew-/*dew-/*du- "to move forward"; *deus-/*dous-/*dus- "substandard, low-quality; to lack (to be pulled under)"; PAA *t?aw- "to submerge"; Arabic in tâkha (t-w-kh) "to dip into a soft substance"; in tâ¿a (t-w-¿) "to dip bread in [butter]"; S du₂₃ **"to submerge (the sign for 'water' has a reading of du₂₃, and, among its meanings is 'inundation')[?] (source: PL *T?E-FA[-SO-])

DISCUSSION: Bomhard has another entry which, as we will see, when amended, is structured very similarly.

There are several problems here. For example, I am not aware of a Sumerian du-ri- meaning 'long time'; there is da-ri₂- with that meaning (and, let us be fair, da does have a recorded reading of du₂₀). Perhaps that is what Bomhard meant? Actually, based on da . . . ri, 'pay off a delivery', which can be interpreted as 'conclude, bring to an end', da-ri₂ is probably better regarded as 'come to a side = end', and better interpreted as 'eternal' rather than just 'long time' except as a deliberate exaggeration. Bomhard's "*t'ow(-A)-/*t'u(-A)- "to leave, to go far away"", I presume is supposed to represent Pokorny's (3.) *dew-, which he defines as ""1. to move one's self forward spatially; to press forward, to distance one's self; therefrom, later 2. temporal prolongation". Bomhard's old-fashioned "A" is presumably an 'a-coloring laryngeal', now usually indicated as H₂, which accounts for the additional reconstructions, recognized by Pokorny, of *dwa:-, *dew6-, and *du:-.

For this, Bomhard offers Arabic "TâHa (base TwH) "to perish, to die; to go away, to depart, to lose one's way, to go astray, to stray, to wander about; to fall, to throw, to cast, to fling, to hurl, to toss, to carry away, to sweep away"". If our proposal is correct, Arabic T should correspond to Nostratic tho; it must, therefore, be rejected.

An even more serious problem is that the IE cognate bears no trace of a meaning 'perish' — quite the contrary, if anything, the idea associated with it is 'prolonged life'. Now, if we had no suitable Arabic cognate, the semantics of this root might be disputable but the Arabic cognates which we will identify show that the basic idea of this root, clearly seen in most of the IE derivations, is 'prolong'.

Forms built on IE *a:/*6 are a reflection not of a 'laryngeal' but a rare sporadic retention of a lengthened vowel caused by previous aspiration of the consonant which preceded it (PL FHA).

We therefore reconstruct PL *T?E-*FHA-, 'heel'+'move-repeatedly'='pull'.

For reasons we probably will never certainly know, another root has been conflated with the IE root in #139 above; and its presence accounts for the meanings of spatial and temporal prolongation. We can only speculate that **dhwa:- was simplified to *dwa:- but, though rare, initial *dhwV- has been reconstructed in four other IE roots. Although these roots have been conflated in IE, in the other languages in which we find a cognate, they are clearly relatable to **dhwa:- rather than **dwa:-.

We therefore reconstruct PL *T?SA-*FHA-, 'long'+'move-repeatedly'='last long, go far, go deep'.

Accordingly, we amend to include:

(Addition AK)PN *t?sawa:- "to last long, to go far, to go deep" > PIE (3.) *dew-/*dow-/*du-; *dhewa:-/*dhw6:- "to move one's self forward spatially; to press forward ", temporal prolongation, "to last long, to go far, to go deep"; in *dheu-b- (for **dhewa:b-) "deep, hollow"; PAA *daw- "to last long"; Arabic in dâra (d-w-r) "to elapse (time)"; in dâla (d-w-l; this is IE *dhwel-, from which English "dwell") "to elapse (time)"; in dâma (d-w-m) "to last, to continue"; PA *Zop'e "hollow under knee or neck"; S zu (for zû) "deep"; su (for **zû_x) "distant, remote; long (duration)" (source: PL *T?SA-FHA)

DISCUSSION: There are a number of interesting roots which have to do with the 'hand', and actions performed with it.

It is amazing to me that Bomhard here again demonstrates the idea that there is some connection between retroflex articulation (PD *r..) and palatality (PN *r^y). Actually, Dravidian retroflex *r.. indicates Nostratic ro.

Rather frequently, when we find a manual activity or action, the initial element is PL *T?A, 'hand', or *T?SE, 'finger'. In this case, because of IE *dregh-, we opt for T?A.

We therefore reconstruct PL *T?A-*RO-, 'hand'+'raise'='hold'.

The Afroasian cognates are rather poorly represented for this root. Egyptologists have identified an Egyptian word Dr.t, 'hand, handle'. This word is also occasionally written with d; and rather than just the result of a well-known confusion in later stages of Egyptian between d and D in some words, I believe it represents a separate root.

Accordingly, we amend:

(124)PN *t?ar^w- "to hold" > PIE (1.) *der-/*dor-/*dR- "(span of the )hand"; PAA *tar- "to hold"; Arabic in tari¿a "to be filled (with = "to hold"); Egyptian in Dr.t (for **dr.t) "hand, handle"; PD *tar..- "to clasp, to embrace"; S in dal-dal "to value highly ("to hold {on to/dear}")" (source: PL *T?A-RO)

However, there are Pyramidic spellings which strongly suggest another root of the form Dr.t; and it appears to have a very similar meaning, perhaps 'hand(ful)', corresponding to IE (2.) *dher-, 'hold, hold firmly, support'.

We therefore reconstruct PL *T?SE-*RO-, 'finger'+'raise'='hold (in a cupped hand or with clenched fingers)'. A further development of meaning from this root seems to be 'prevent (by holding tightly)'.

We have proposed, above, a number of examples of the correspondence Nostratic *t?s = Sumerian z = Dravidian *Z and possibly = Dravidian *c = IE *dh = Egyptian D (but also ') = Arabic d (but also D).But there are also a few roots in which Sumerian t = Dravidian *t and possibly = Dravidian *tj = IE *dh = Egyptian D = Arabic d. This seems to be indicating that a PL *T?A-*¿E-, 'hand-like'='finger', was also in use as a synonym for *T?SE, 'finger'. And the correspondences suggest that PL *T?A-¿E became Nostratic *t?ey for one group of languages (Dravidian, Sumerian) but became *t?se for another (Arabic, Egyptian, IE).

In spite of the reconstructed alveolar *r, which would normally indicate Nostratic *r(h)e, we will include Dravidian *tje:r-, 'handful', and *ti:r- 'be finished', on the questionable assumption that it represents a fuller form (*t?ayeroy- ), possibly dimly reflected in IE *dher6-, a companion form to IE (2.) *dher-.

Now Bomhard does have an entry, in which he appears to be attempting to analyze this root:

(143)PN *d^yar-/*d^y6r- "to hold firmly" > PIE *dher-/*dhor-/*dhR- "to hold firmly in the hand, to support"; PAA *d^yar-/*d^y6r- "to hold firmly; hand, arm"

Bomhard apparently believes that palatalization will result in aspiration, an idea which is dubious at best. In another publication, he cites Arabic dhirâ¿-un, 'forearm', as a cognate; and I am inclined to accept this as a possible derivation from Nostratic *t?ser^w- listed below.

Accordingly, we amend to include:

(143)PN *t?ayer^w- (also *t?ser^w-) "to hold (in a cupped hand or with clenched fingers = to hold tightly)" > PIE (2.) *dher-/*dhor-/*dhR- "to hold, to hold firmly, to support, to hold back (from)"; PAA *dar- "to hold"; Arabic in darija "to keep to (a religion)"; in darasa "to disappear"; perhaps in dhirâ¿-un "forearm" (from *t?ser^w-)[?] ; Egyptian in Dr.j "retaining wall"; Dr "to end, to hinder, to obstruct"; S til "to put an end to, to finish, to cease"; PD *tje:r- "handful"; *ti:r- "to be finished" (source: PL *T?A-¿E-RO and *T?SE-RO)

There is yet another root which can be reconstructed for Nostratic that has a similar semantic range.

Here, the underlying PL form appears to be *T?SE-*FA-*RO, 'finger'-set-'raise', again something like 'cup the hand', 'handful', 'hold' = 'contain'. The IE root *der-, discussed above as amended #124, includes forms that Pokorny reconstructs as *dwer-. Similarly to what we proposed above, we believe that an IE **dhewer has here again been simplified to *d(e)wer-.

Supporting this line of thought is that the Egyptian sign currently read as Dr, which pictures a 'bundle of flax-stems, showing the boles', discussed above as amended #124, includes forms that correspond to Pokorny's reconstructions of *dwer-, listed as an alternate form under *der (e.g. in nDrj, 'grasp, hold fast, catch, arrest, take possession of, observe (regulations), hold to (orders), follow (path), draw tight (lips of wound), imprison, suppress'); but also to forms which correspond to derivatives listed under IE root (3.) *dew-, 'move one's self forward spatially; press forward', temporal prolongation, 'last long, go far, go deep' (e.g. Dr, 'since, before, until [lasting to]'; in sDr, 'spend the night'; and, in Dr.j, 'strong [= durable, lasting]'), discussed in Addition AK above; and this is the basis for an extended form: Nostratic t?sawa:r^w. Similarly to what we proposed above, we believe that an IE **dhewa:-ro has here again been simplified to *dwa:-ro-.

Accordingly, we amend to include:

(Addition AL)PN *t?sewar^w- "to hold, to contain" > PIE (1.) *dewer-/*dower-/*dur- (from **dhewer) "(span of the )hand"; PAA *dawar- "to hold"; Arabic dâwara "to control, to look steadily at"; in ?adâra "to manage, to avert"; Egyptian in Dr.j (for **Dwr.j) "retaining wall"; Dr (for **Dwr) "to end, to hinder, to obstruct"; in n(-)Dr(j) (for **nDwr.j) "to grasp, to hold fast, to catch, to arrest, to take possession of, to observe (regulations), to hold to (orders), to follow (path), to draw tight (lips of wound), to imprison, to suppress"; S **zul_x (the sign reading zal, and accorded the following meanings, is also read sul₂, which, based on the recognized confusion between Sumerian s and z, I suggest should be read **zul_x; zal is a result of Nostratic *t?sar^w [PL *T?SA-*FHA-*RO, 'long-move repeatedly-very'-'last long', and extension of the root tread in Addition AK above] ) "supply (=contents)"; "to be full (of" = "to contain")" (source: PL *T?SE-FA-RO)

I have long maintained that Egyptian neutralized vowel-quality distinctions, and thus had a phonology very similar to Old Indian. This fact, if it is one, that Nostratic *t?sewar^w- and *t?sawa:r^w- could be written with the same sign in Egyptian, suggests strongly that the underlying form was [dzawa(:)r] for both; and that vowel-quality distinctions had been neutralized (note, however, that a partial distinction in Egyptian among former vowel-qualities was maintained by most obstrucents having two forms: one derived from Nostratic Ca/e (Egyptian 3) and another derived from Nostratic Co (Egyptian r).

Here, the underlying PL form appears to be *T?SA-*FHA-*RO, 'long-move repeatedly-very', again something like 'last long'.

And we further amend to include:

(Addition AM)PN *t?sa(wa:)r^w- "to be (last) very long(, to go very far, to go very deep)" > PIE *dewro-/*dowro-/*duro- (for *dhewa:-ro-/*dhw6-ro-) "long, far distant"; PAA *dawar- "to last long"; Arabic dâra (d-w-r) "to elapse (time)"; Egyptian Dr (for **Dwr) "since, before, until"; in s(-)Dr "to spend the night"; in Dr.j "strong (= durable, lasting)"; S **zal ( for **zul_x; the sign reading zal, is also read sul₂, which, based on the recognized confusion between Sumerian s and z, I suggest should be read **zul_x; I presume zal derives from an alternate form without the medial *FHA, meaning simply "very long") "to flow; to continue; to pass, to elapse (said about time); to spend the day; to tarry, to wait" (source: PL *T?SA[-FHA]-RO)

As a corollary of the my proposal that Egyptian obstruents have two forms based on the vowel-quality of the following vowel is the idea that apparent voice distinctions in obstruents are actually distinctions in the following vowels, and that voice distinctions were neutralized in Egyptian; therefore, Egyptian D represents both Nostratic *t?sa/e and *thsa/e while Egyptian ' reflects Nostratic *t?so and *thso. We have been investigating roots in which Egyptian D represents IE *dh; we will now investigate a root in which Egyptian D represents IE *t, i.e. *th.

(149)PN *t^y[h]ar-/*t^y[h]6r- "to advance toward an end or goal; to attain or achieve an end or a goal, to reach, to come to, to arrive at; to master, to become master of" > PIE *t^[h]er-/*t^[h]or-/*t^[h]R- , (extended forms) *t^[h]erhH-/*t^[h]orhH-/*t^[h]RhH-, *t^[h]rehH- [*t^[h]rahH-] /*t^[h]rohH- (> *t^[h]ra:- /*t^[h]ro:-] "to advance to or toward an end or goal, to pass across or over, to pass through; to attain or achieve an end or a goal, to reach, to come to, to arrive at; to overcome, to master, to become master of, to control"; PAA *t^y[h]ar-/*t^y[h]6r- "to advance to or toward an end or goal; to attain or achieve an end or a goal, to reach, to come to, to arrive at; PD *ca:r- "to reach, to approach, to be near to"; S šár "to bring together", reduplicated šár-šár "to arrange in order, to set or put in order, to organize"

In another publication, the only "PAA" cognate Bomhard can offer is Egyptian TrTr, 'overcome, destroy'. Again, Bomhard has utilized an entry from Budge's dictionary, which is unknown in this meaning to Erman-Grapow. It is certainly not attested in Old or Middle Egyptian; and, in all likelihood, is a Semitic loanword with the meaning 'siege-wall' and 'besiege'. To base a PAA reconstruction on a word like this, late, rare and of unknown provenance and uncertain meaning, in only Egyptian, is reprehensibly irresponsible and intellectually dishonest. Frankly, I had no idea Bomhard could be capable of such "scholarship", and I am deeply disappointed.

In any case, his "cognate" is immaterial since Egyptian T (underlined t) can only represent IE *g or *k, and hence cannot be the first consonant of a word cognate with IE *ter-.

Similarly, the proposed Sumerian "cognate" is impossible; Sumerian š represents Nostratic *s(h)e and *x(h)e, and cannot be the first consonant of a word cognate with IE *ter-. But worse yet, šar₂ is unknown in the meaning 'bring together'; and it appears that Bomhard has invented it out of 'be many, multiply, mix', its normal meaning. And worst, 'bring together' does not seem to be the basic meaning of this root: instead, it is 'cross over' [IE (4.) *ter-] or 'go through' [IE (3.) *ter-].

Once we realize this, it can be seen that the meanings for the IE root (4.) *ter- can easily be unified under the idea of 'cross over' ("to advance to or toward an end or goal, to pass across or over, to pass through; to attain or achieve an end or a goal, to reach, to come to, to arrive at; to overcome, to master, to become master of, to control"). And, additionally, the Arabic equivalent can also be easily found: dharra, 'rise (sun)' = 'cross over the horizon'; dharrafa, 'exceed' = 'cross over an amount'; tadharrâ, 'ascend to the top of' = 'cross over the top'. In addition, the Egyptian cognate is also easily found: D3.j, 'ferry across, cross'; D3.t, 'remainder'.

More problematical is the Sumerian root. The normal 'cross over' in Sumerian is dib_(1,2). The word dib₂ is written with a sign (Jaritz #893) which, when combined with a sign reading zib₄ (Jaritz #15), reads zir₂. This is, perhaps, a rather unusual way of suggesting that zir₂, which, so far as I know, has no established meaning, should be equated with dib₂ (defined in Halloran's Glossary as "to pass (by/along); to send over; to traverse, cross; to wander") in the meaning 'go through', corresponding to Dravidian *cer-, 'put in, insert'. And Sumerian dib_(1,2) is probably cognate with IE *(s)te/e:ip-, 'puncture' (cf. dib, 'to fish'). And another Egyptian D3j, which pictures a 'fire-drill', is acknowledged to mean 'pierce, transfix'.

A Sumerian sign reading zar means 'remainder'. By itself, perhaps meaningless, but combined with Dravidian *ca:r- (which, let me hasten to add, is not reconstructed at Starostin's website), it suggests that the Nostratic form of the root is question was *thsar^y, 'go/cross over', and that there was a second root *thser^y, 'go through'. If this root is in Bomhard's dictionary, I have been unable to find it.

So, we reconstruct *THSA-*RHE, 'rear-up-come-down' = 'come over'.

And we amend:

(149)PN *thsar^y- "to come over/across" > PIE (4.) *ter-/*tor-/*tR- (for *ther-/*thor-/*thR-) "to reach the other side, to press through, to cross over, to overcome, to catch up with, to bring across, to save"; PAA *dhar- "to come across"; Arabic dharra "to rise (sun)" = "to cross over the horizon"; dhara¿a "to overcome someone (vomit)"; dharrafa "to exceed" = "to cross over an amount"; tadharrâ "to ascend to the top of" = "to cross over the top"; Egyptian D3j "to ferry across, to cross"; PD *ca:r- "to reach, to approach, to be near to"; S **zar "remainder" (source: PL *THSA-RHE)

So, we further reconstruct PL *THSE-*RE, 'bristle-out-apply' = 'cause to extrude, pierce'.

And we amend to include:

(Addition AN)PN *thser^y- "to cause to extrude, pierce" > PIE (3.) *ter-/*tor-/*tR- (for *ther-/*thor-/*thR-) "to bore through"; PAA *dhar- "to pierce"; Arabic dharra "to come forth (horn, herbage)" = "to pierce skin, surface"; dhariba "to be sharp" = "to be capable of puncturing"; Egyptian D3j "to pierce, to transfix"; PD *cer- "to put in, to insert (better, "to puncture")"; S **zir₂ "**to cause to go through" (source: PL *THSE-RE)

DISCUSSION: An important root which Bomhard seems to have completely overlooked will illustrate the correspondences of Nostratic *th.

To begin with, we have IE *ste:/er-, 'star'. The sign which depicts a 'star' in Sumerian has the reading dir₃. The sign which is a 'star in a circle' in Egyptian reads d3. Finally, we have Dravidian *teri-, 'clear, evident, bright, shine'. There is also Altaic *t'iri-, 'morning', which fairly probably is to be included as well. In spite of the final *i of Dravidian *teri-, the tap-r suggests Nostratic *ra, and variants of IE *ste:/er- (stera:-) suggest Nostratic *ra:, with the sporadic lengthened *e: a result not of a 'laryngeal' or a former affricate but rather as compensation for lost *a:. The initial *s of the IE form is s-mobile, and not an essential part of the root, as forms like *te:ra:- show clearly.

The predicted Arabic form would be **thar- and, in the meaning of 'star', cannot be found. However, there is the curious name for the Pleiades, thuraiyâ, which also designates a 'cluster of lamps' and 'luster'. Could this possibly be an extension of this missing root?

All these words, with the exception of the Altaic form, can be easily related under one Nostratic form: *thera:-, and be reconstructed as: PL *THE-*RHA, 'shine-bird' = 'star'.

So we amend to include:

(Addition AO)PN *thera:- "star" > PIE (2.) *ste/e:r- (for **tera:-) "star"; PAA **thar- "**star"; Arabic in thuraiyâ "Pleiades, cluster of lamps, luster[?]"; Egyptian d3 "**star"; PA probably in *t'iri- "morning"; PD *teri- "clear, evident, bright, shine" (*thera:-ye- "star-like"); S dir₃ "**star" (source: PL *THE-RHA)

(99)PN *tar-/*t6r- "to be dry, arid" > PIE *t^[h]ers-/*t^[h]ors-/*t^[h]Rs- "to be or become dry"; PAA *tar-/*t6r- "to be dry, arid"

DISCUSSION: One of the principal methods of 'drying' is 'heating'; and I believe, semantically, the idea of 'heating' is primary in this root.

This line of inquiry is strengthened by the fact that Egyptian t3 means 'earth' but also 'be hot, kiln'; as we have mentioned, Egyptian t corresponds to Nostratic t?/ho. In IE, we have *ters-, 'dry, dry out; thirst, be thirsty'. This narrows the Nostratic reconstruction to *tho for the first element. In addition there is Sumerian dur₂, 'dry out'; and dur(-)una₂, 'oven for baking and roasting'; this confirms the *o-quality of the first element. It is appropriate at this point to explain an exception to the correspondences I have earlier explained for Sumerian.

Generally, all Nostratic unaffricated stops are transcribed in Sumerian as voiced stops: thus *p?, *t?, *k?, *ph, *th, *kh all appear as Sumerian b, d, g. Four affricated stops of Nostratic, *p?f, phf, k?x, khx show up in Sumerian as voiceless stops: p, k. However, in the apical series, Nostratic t?s, ths appear in Sumerian as z (presumably [ts]). But these rules have an exception: Nostratic t?o, tho, t?so, thso all are written with Sumerian tu; and where cuneiform makes a distinction, also with Tu, i.e. subdotted-t., and perhaps also Su, i.e. subdotted-s.. Presumably, this is due to a retroflex pronunciation, which, as Arabic demonstrates, is most tenaciously retained in the apical stops and affricates.

The currently accepted reading of dur₂ for 'dry out' should therefore be amended to tur₇. Currently, dur₂ is an alternate reading for duruna. I presume that duruna is equivalent to duruna₂; and therefore, I consider tur₇ to also have the meaning 'oven for roasting or baking', analogous to Egyptian t3.

In addition, there are Altaic *t'ó:ré-, 'soil, dust'; and Dravidian *tur-, 'dust, dirt, straw'; both reconfirming the apical onset and vowel-quality of the first element; the alveolar *r of the Dravidian suggests a final element of Nostratic *r(h)e.

On the other hand, in another publication, Bomhard cites Arabic tariba, 'become dusty'; turbâ, 'dust, earth'; and taraza, 'be hard, dry, arid'; as derivations from this root. The proposals we have made predict Arabic **Tar-; and I am at a loss to explain the discrepancy since I consider it very likely that these words are directly or indirectly derived. Strengthening this conclusion is the existence of IE *(s)terbh-, 'dry', corresponding substantially with Arabic t-r-b, which is probably also a derivation from this root. I know of no convincing examples of 'emphasis' (retroflex articulation) being lost; and so I will simply include these words with a question mark.

The IE form *ters- is, in my opinion, a result of the Nostratic root analyzed below + Nostratic *sa, 'strong' (PL *SA), to which Arabic taraza, 'be hard, dry, arid', may be compared (overlooking the problem with initial t).

All these words, with the exception of the Arabic forms, can be easily related under one Nostratic form: *thor^y-, and be reconstructed as: PL *THO-*RE, 'heat-apply' = 'heat, dry out'.

So we amend to include:

(99)PN *thor^y- "to heat, to dry out; dust" > PIE *ters-/*tors-/*tRs- "dry, to dry out; thirst, to be thirsty"; PAA **Tar- "to heat, to dry out; dust"; Arabic in tariba (for **Tariba[?]) "to become dusty" [cf. IE *(s)terbh- "dry"]; turbâ "dust, earth"; taraza "to be hard, dry, arid"[?]'; Egyptian t3 "earth, be hot, kiln"; PA probably *t'ó:ré- "soil, dust"; PD *tur- "dust, dirt, straw"; S **tur₇ "to dry out, **oven for baking and roasting" (tur₇ is another reading of the sign for dur₂, the currently accepted reading for this meaning) (source: PL *THO-RE)

DISCUSSION: Bomhard has three roots (two entries) which I believe are resolvable into two entries.

(102)PN *t^[h]i / *t^[h]e "you" > PIE (nom. sg.) *t^[h]u/u: "you"; (acc. sg.) *t^[h]we/e: / *t^[h]e/e:; (2nd pl. verb ending) *-t^[h]e; PAA *t^[h]a / *t^[h]6 "you"; PU (sg.) *te "you", (pl.) *te "you"; Elamo-Dravidian: Elamite (2nd sg. verb ending) -t; Dravidian: Parji (appositional marker of 2nd sg. in pronominalized nouns and verb suffix of 2nd sg.) -t; PA *ti (> CM *chi); S za-e "you", 2nd sg. possessive suffix -zu "your"; Eskimo-Aleut: West Greenlandic 2nd sg. absolutive possessive suffix -(i)t

and

(103)PN *t^[h]a-/*t^[h]6- "this", *t^[h]u-/*t^[h]o- "that" > PIE *t^[h]o- "this, that"; PAA *t^[h]a-/*t^[h]6- "this, that"; PU *ta / *tä "this"; *to- "that"; PD *ta/a:(m)- "they, themselves"; PA *te "that"

I believe the key to reconstructing the Nostratic root for 'you' (singular) is the Sumerian word za, 'you' (singular). If the correspondences we have been advocating are correct, Sumerian z will reflect Nostratic *t?sa or *thsa. If IE *te, which I take to be the basal form, is related, the Nostratic antecedent can only be *thsa. Pokorny does not reconstruct *t^([h])e:; and I believe this entry of Bomhard is in error. If it were not, we should be forced to explain the retention of Nostratic **e through sporadic lengthening with the loss of aspiration from *th. With *te, we have no such problem: Nostratic *thsa becomes IE *the/o, and *te. However, we do have a lengthened form in *ta:, the "feminine" of *to, 'this'. I believe this is the direct reflex of Nostratic *tsha, and that *to is a back-formation based on the incorrect analysis of *ta: as a feminine; however, it is also possible that *ta: is a result of Nostratic *tshah (PL *THSA-*HA, 'alone'-feminine = 'this (one) [feminine]').

IE *tu/u: is a result of Nostratic *thsa, 'this (one)' + the topical (circumlocative) suffix *w (PL *FA).

In addition, Nostratic *tsha, 'this (one), you', shows up in a number of verbal inflections as IE *th in Old Indian, the only language family that seems to sporadically preserve the earliest IE reflex: in 2nd person dual primary *-t(h)e/os (for *-thas and *-thawas; in 2nd person plural primary -tha; in 2nd person singular middle imperfect secondary -tha:H (not IE -*the:s but -*tha:s); in 2nd person dual middle present -a:the: for *-V:thai; in 2nd person plural middle primary -dhve: and secondary -dhvam (for -*th(a)wai and -*th(a)wam with *th being voiced before *w; in 2nd person dual imperfect middle thematic -e:tha:m (for -*Vitha:m); and also, preserving the vowel-quality and length: in 2nd person dual secondary -*ta: (Baltic and Slavic), for -*tha.

The predicted reflex of Nostratic *tsha/e in Arabic is dh; and Arabic hadhâ, 'this', shows it as predicted combined with another element (ha-). The PAA 2nd person formant *t cannot be a result of the amended root for 'this', *tsha nor yet Bomhard's *the, which would result in Arabic **tha.

In fact, in spite of the similarity eagerly embraced by other Nostraticists, it is questionable to me whether Arabic t is, at origin, an element of a 2nd person formant corresponding to IE *t. If it were, what is it doing in the 1st person singular perfect: katabtu, 'I read', and in the feminine 3rd person singular katabat? I propose that Arabic t in the first and second person, and third person feminine, is derived from Nostratic *t?a, 'side, with (comitative)', and simply designates any person close by, literally or figuratively; and, as such, is unrelated to IE *t(h) as a component of designations for the 2nd person. This is a Semitic innovation, and cannot be projected back into Nostratic. The particle tâ, used with oaths, seems to mean roughly 'by, with'; and together with tâ, 'this', suggests strongly to me that Semitic *t, is derived from Nostratic *t?a; and is merely a Semitic substitute for Nostratic *tsha as a paraphrase for the 2nd person. However, although associated with the first person, we should not overlook IE *de/o-, a demonstrative stem, which possibly reflects Nostratic *t?a, and could have been cognate with the Semitic *t discussed above.

Bomhard has this root (almost) with his entry:

(146)PN *d^yi-/*d^ye- demonstrative stem > PAA *d^ya-/*d^y6- demonstrative stem; PU *t^yi / *t^ye demonstrative stem: "this, that"

In another publication, Bomhard reveals that, in his opinion, this is the source of Arabic dhâ, 'this'. We have shown, in a number of examples, that it is Nostratic *tsha/e that leads to Arabic dh. And, in that same publication, Bomhard includes PIE "*-d^[h]e" which he omits from the entry in the present dictionary, which was advised.

So far as I am able to determine, Bomhard has never accounted for Arabic tâ or IE *de/o-; and I believe that this is the proper entry in which to account for them.

As for "PU *t^yi / *t^ye", Rédei does not have it. If, in fact, it can be shown to exist, it probably would be a derivation of Nostratic *t?ay.

We reconstruct: PL *T?A-(*¿E-), 'with'(+'like'='this, he/she').

No Dravidian cognate can presently be identified.

Accordingly, we amend:

(146)PN *t?a-(y) "nearby, with(, this)" > PIE *de-/*do- "here, this"; *dey-/*doy-/*di "he, she"; PAA *ta(y) "nearby, with(, this)"; in Arabic tâ (**tay[?]), used with oaths: "by, with"; in tâ (**tay[?]) "this"; probably -t in 1st and 2nd person inflections of the imperfect, and 2nd person of the perfect; in Egyptian dy "here"; PU *-ta "with" in Vogul -t, locative; *t^yi[?] (for **tay[?]) "this"; PA *-da locative; S -da "with", inflectional suffix (source: PL *T?A[-¿E])

Bomhard cites "PU (sg.) *te "you", (pl.) *te "you"" but Rédei writes: "Der Stammvokal kann nicht auf einen einzigen FU Vorgänger zurückgeführt werden". I believe this is because a Uralic **ta, seen in *ta, 'that one here, this one there', has been combined with *y, producing *te (cf. IE *tyo-) and *w producing *to (cf. IE *tew-), in order to rather imperfectly differentiate uses of 'this': nominal ('this-one' = *ta), pronominal ('you' = *to), and pronominal demonstrative ('this' = *te). In view of the fact that Rédei's reconstruction for Uralic 'this' shows a similar alternation, I conclude that the original form underlying both is **ta. In view of the lengthened vowel of Finnish taa, it is possible that *ta, 'that one here, this one there', is really **ta:, and a product of *tsha + *?a, 'here' (PL *THSA-*?A, 'alone-here' = 'this (one) here'), seen again in the 1st person singular prefix of the Semitic perfect conjugation (?a(-)ktabu).

Starostin's website gives *ná for Altaic 'thou'; *ko:, 'this'; and *ó, 'this, that'. Starostin does has Mongolian *te-re, 'that'; Tungus *ta, 'that'; and Korean *tjá, 'that'. This seems rather a poor basis for Bomhard's "PA *ti", 'you', which is based, according to another publication on Poppe (1960). I reject this reconstruction as outdated, not sufficiently substantiated, and probably incorrect.

Why Bomhard cites the ergative form (Sumerian za-e) rather than the absolutive (za) , which is regularly formed by the absolutive case + e, will remain forever Bomhard's secret.

While there are forms in individual languages for 'that' suggest the possibility of an Altaic reconstruction containing initial *t, the data is insufficient, in my opinion, to justify a reconstruction of "PA *te "that""; and it will be omitted.

As for PD *ta/a:(m)- , 'they, themselves', if the vocalic correspondences I (and I presume, Bomhard) have been proposing have any validity, then, rather than with 'that' as Bomhard has it, this form would go with 'this'.

The cognates for 'this, you', can be easily related under one Nostratic form: *thsa-, and be reconstructed as: PL *THSA-, 'alone' = 'this, you (singular)'.

The nature of the first element of 'that' is provided by Egyptian '3, 'there', which can only reflect Nostratic *t?so or *thso. IE *te/o- for 'that' decides the issue in favor of Nostratic *thso; and Uralic *to-, 'that', confirms the vowel-quality.

The cognates for 'that' can therefore be easily related under one Nostratic form: *thso-, and be reconstructed as: PL *THSO-, 'around' = 'that'.

I will confess that I am tempted to think, in the basis of the semantics of *thsa, 'alone', and *thso, 'around', that PL *THSO may have originally been a suppletive plural for *THSA. Supporting that speculation minimally is the IE suffix -tV/V:-, which, as most readers will know, forms abstract substantives; and, which might have started life as -*thV. Additionally, it may be of interest that Tibetan forms a plural in -tsho.

Accordingly, we amend to include:

(102)PN *thsa(w-/y-) "this (one), you" > PIE *te- (for **the/o) "this", *te- (for **the/o) "you"; *ta: (for earlier **tha if not from *thaH [Nostratic *tshah]) "this (one), you" (later reinterpreted as feminine); in *tew-/*tow-*tu (from **the/ow- for earlier **thaw- [Nostratic *tshaw]; cf. IE *ta:wont "of such a size") "you (nominative, originally circumlocative-topical); Old Indian in 2nd person dual primary *-t(h)e/os (for *-thas and *-thawas; in 2nd person plural primary -tha; in 2nd person singular middle imperfect secondary -tha:H (not IE -*the:s but -*tha:s); in 2nd person dual middle present -a:the: for *-V:thai; in 2nd person plural middle primary -dhve: and secondary -dhvam (for -*th(a)wai and -*th(a)wam with *th being voiced before *w; in 2nd person dual imperfect middle thematic -e:tha:m (for -*Vitha:m); and also, preserving the vowel-quality and length: in 2nd person dual secondary -*ta: (Baltic and Slavic), for -*tha; PAA *dhay in Arabic [ha(-)]dhâ "this"; PU *ta "this one"; *te "this" (*thsay); *to "you" (*thsaw); PD possibly *ta/a:(m)- "they, themselves"; Elamo-Dravidian: possibly Elamite (2nd person singular verbal ending) -t; Dravidian: possibly Parji (2nd person singular verbal ending) -t; S za "you", in 2nd singular. possessive suffix -zu (for -zû from *-thsaw) "your"; Eskimo-Aleut: West Greenlandic 2nd person singular absolutive possessive suffix -(i)t (source: PL *THSA[-FA/-¿E])

and

(103)PN *thso- "that (one)" > PIE *te/o- (for **the/o-) "that"; in *tor-/te:r- "there"; PAA **Za "that"; Egyptian in '3 "there"; PU *to- "that" (source: PL *THSO)

DISCUSSION: We have briefly discussed this root earlier in connection with Bomhard's entry #145.

The Arabic word dhawâ, 'wither, wilt, fade', is analyzable as: PL *THSA-*FHA, 'elongate-itself-move-repeatedly' = 'disintegrate, melt'.

(Addition AP)PN *thsawa: "to fall apart, disintegrate, to melt" > PIE *ta:u- (for *thau-) "to melt, to rot, disintegrate"; PAA *dhaw "to rot, to disintegrate"; Arabic in dhawâ "to wither, to wilt, to fade"; in dhâba (dh-w-b; cf. IE *ta:bh- [for **thabh-] "gradual disintegration") "to melt, to dissolve, to liquify"; PD in *Cab- "rotten"; S zu₄ "**to rot" (a sign which reads sugin, and has the meaning "rot, decay", also reads zu₄) (source: PL *THSA-FHA)

DISCUSSION: An animal name over which much ink has been inordinately spilt is the name for the 'ox', seen in Latin as taurus, which can be referred back to IE *ta/eur-. The Arabic form, thaur-un, assures us that the IE *t represents Nostratic tha/e. Dravidian *tor-, 'kine', informs us through the alveolar r that the final Nostratic vowel was *e. The sporadic retention of *a in the IE form is, I believe due to the occasional lengthening effect of a following syllable that was, itself, long; and, on the basis of the data above, I can reconstruct Nostratic thawa:r^y-.

Nostratic thawa:r^y- would come from PL *THA-*FHA-*RE, 'damp-move repeatedly-apply' = 'dampen'.

The anticipated Sumerian form of Nostratic thawa:r^y- would be **dûr, and dur₂-dur₂, 'dam to create a reservoir', certainly seems like the correct semantic range, since the effect is to 'dampen' or 'irrigate' the ground of the reservoir. In addition, Sumerian has duru₅, 'irrigated, moist, fresh'.

Dravidian *tor-, 'kine', once again illustrates that Nostratic *Vw appears as *o in Dravidian. And because it does, we can consider Dravidian *to:-, 'be wet', to represent the first two elements of the word. We would anticipate Sumerian **dû, and, although we have no assigned meaning for it, du₂₃ is a reading for the Sumerian sign which means 'water', which suggests strongly the possibility that du₂₃ means, if not 'dew', and least 'dampness'. This same sign reads dur₅ in the meaning 'anoint', which is, of course, also a kind of 'dampening'. And finally, probably to be put here also, is Dravidian *to:r-, 'flow (blood)'.

Nostratic *thawa:- would appear in IE as *teu- (possibly also as *ta:u-), and IE *teu- is reconstructed for the meaning 'bedew, dampen'.

For Arabic, we have thâba (th-w-b), 'collect (water)'; and thâ¿a (th-w-¿), 'flow (water)', which establish th-w as a basis for words associated with 'dampness'.

So, it appears, according to thawa:r^y-, that PL speakers characterized the 'ox' as the 'dampener', presumably referring to the capacity of its urination.. I would be skeptical of where the data had led me but for the fact that oxen are domesticated varieties of an extinct animal that is called the 'aurochs'. The second element of 'aur-ochs' is provided by IE *weg^w, 'damp, moisten', to which Egyptian jwH, 'moisten', can be compared; the first element is referred to IE *awer-, 'water'.

Although cognates cannot presently be located in Altaic and Uralic, there is a possibility that an Egyptian cognate may be identified. Nostratic thawa:r^y- would appear in Egyptian as **dw3. There is a Late Egyptian attestation of dw3.yt as 'garden', which might be interpreted as 'irrigated plot'. And a Late Egyptian name for the Hippopotamus Goddess, associated with the inundation, is recorded as dw3.t. Also, dw3 is attested in the Pyramid Texts as a 'body of water in heaven'; and associated with dw, a verb of undetermined meaning. Finally, dw3 is the common word for 'morning', the time of 'dews'. I think these data indicate, at least, the possibility that dw(3) might have been associated with 'dampness'.

Bomhard does offer an entry for 'bull, steer', in #148, which he reconstructs as *t^y[h]awr- but then on page 234 informs us that "the etymology should be removed — we are probably dealing with loanwords here". I disagree. We have been able to find a number of cognates which suggest strongly, on the basis of their regular correspondences, that this word is very old and well-distributed among the daughter-languages.

Accordingly, we will amend:

(148)PN *thawa:(r^y) "to dampen (dampener = ox)" > PIE *tew-/*tow-/*tu- "to bedew, to dampen"; in *teur- (also *taur- from *ta:ur-) "ox"; PAA *thaw "to dampen"; Arabic in thâba (th-w-b) "to collect (water)"; in thâ¿a (th-w-¿) "to flow (water)"; in thaur-un "ox"; Egyptian possibly in dw3.yt "garden (if = irrigated plot)"; possibly in dw3.t "name for Hippopotamus Goddess (if = dampener [fem.])"; possibly in dw3 "body of water in heaven (if = damp place)"; possibly in dw3 "morning (if = [time of] dew)"; PD *to:- "to be wet"; in *tor- "kine"; possibly in *to:r- "to flow (blood)"; S du₂₃ (for dû₂₃) "**dew, dampness"; in dur₂-dur₂ (for **dûr-dûr) "dam to create a reservoir"; in dur₅ (for dûr₅) "to anoint"; duru₅ (for dûru₅) "irrigated, moist, fresh" (source: PL *THA-FHA[RE])

DISCUSSION: Bomhard has made the same mistake as Pokorny in deriving IE (2.) *ter-, 'tender, weak', from (3.) *ter-, 'rub, rub while turning, turn, bore through', because in IE both Nostratic *ro and *re appear as *r. But, with the advantage of Egyptian data, we can see that (3.) *ter- is Egyptian D3 (discussed above under Addition AN, PN *thser^y- "to cause to extrude, pierce"), and (2.) *ter- is Egyptian tr; what is puzzling is why Bomhard should not have noticed this. In addition, Nostratic *re or -*r^y remains r in Sumerian but Nostratic *ro or -*r^w appears in Sumerian as l.

(96)PN *t^[h]ar^y-/*t^[h]6r^y- "weak, frail, delicate" (derivative of the preceding) > PIE *t^[h]er-/*t^[h]or- "weak, frail, delicate"; Afroasiatic: Egyptian tr "to be weak"

If we approach (2.) *ter- without assuming a derivation from (3.) *ter-, we obtain a very different semantic interpretation. The correct IE root with which to associate it is (1.) *(s)ter-, 'rigid, be stiff, strut'. And in the sense associated with 'young', Latin stre:nuus, 'active, vigorous', points the way. This term, when applied to a fruit, describes firm texture and resilience, characteristics often associated with freshness. It has been reapplied to animals; and here it indicates the vigor of youth as *(s)ter-g-, 'strong', also indicates. In addition, it provides the basis for words meaning 'reed' and 'cane' ('resilient'). If we look at the entry under (2.) *ter-, we find also *tor-no-, 'young creature'; and many derivatives meaning 'young'. And although Pokorny also mentions "schwach", 'weak', based on his imperfect understanding of the relationships involved (perhaps based on 'tender', a characteristic of young animals and plants, is not necessarily 'weak').

Sumerian tur is the reading currently employed for a sign that means 'young'. Differing with Jaritz, I do not believe the archaic sign portrays 'two female breasts' but rather 'two pieces of fruit connected by a branch or vine'. However, the sign also reads Tur₃ and tul₄; in addition, with the reading dumu, it is the common word for 'child'.

Bomhard implies that Egyptian tr, 'be weak', is well-established, but his only source for it is Budge's An Egyptian Hieroglyphic Dictionary, which, as previously mentioned, is a compendium of spellings derived from words throughout Egyptian history; and can be very misleading. The word is not present in Erman-Grapow, the hieroglyphic Egyptian dictionary upon which Bomhard or any serious researcher should primarily rely. However, I will accept the existence of the word (because I have other PAA cognates) but not the meaning Budge assigns it. As Budge has it, the word has a determinative of 'child sitting (on lap) with hand to mouth'. This determinative accompanies rnpj, 'be young', and šrj and Xrd, 'child', but words for 'weak' have another determinative: 'man sinking to ground from fatigue' (Gardiner's A7). According, I amend tr to 'be young'.

The sign used to write tr, which, contra Gardiner, I believe represents a 'palm shoot with bud', is also used as a second determinative for rnpj, 'be young'; as tr, it is probably cognate with IE (7.) *(s)ter-, 'pointed plant shafts'; and appears to also have the reading (j)rj, which may relate to IE *er, 'shoot'. In addition, Albanian shtjerrë, 'young cow, lamb', from (6.) ster-, 'infertile', suggests that the infertility is attributable to lack of maturity.

Egyptian r can only represent Nostratic *ro and *rho. However Nostratic *rho results in IE *l; therefore, the final consonant of the root must be Nostratic *r^w. However, Nostratic *r^w normally results in Sumerian l also, and therefore we amend Sumerian tur to tul₄ for the meaning 'young'.

The Arabic cognates are easily to find once one knows where to look. As mentioned above, Nostratic *tho should appear in Arabic as T; and we find Tarra, 'grow forth newly'; Tarû?a, 'be fresh, juicy'; Tarû (T-r-w), 'be quite fresh'; and Tarîy-un (T-r-w), 'fresh'.

These cognates can therefore be easily related under one Nostratic form: *thor^w-, and be reconstructed as: PL *THO-*RO-(*FO[?]-, 'be-compacted-very'(+'leaf') = 'firm, fresh, young(growth)'.

Let me be the first to point out that the vocalization of the final formant (*FO) is purely a guess. There is no language derived from Nostratic that would allow us to definite assign a vowel-quality to the final syllable.

Accordingly, we amend:

(96)PN *thor^w(/row^(w)-) "firm, fresh, young(growth)" > PIE *ter-/*tor-/*tR- (for **te/o-ró-), *teru- "tender, weak"; PAA *Tar, *Taraw "fresh"; Arabic Tarra "to grow forth newly"; in Tarû?a "to be fresh, juicy"; in Tarû "to be quite fresh"; in Tarîy-un "fresh"; Egyptian tr "**to be young, fresh"; PA *t'ó:rV "young animal"; PD possibly *to:r.- "friend, younger brother, assistance, help (if = tender{ness})[?]"; S tul₄ (for currently read tur) "young" (source: PL *THO-RO[-FO{?}])

DISCUSSION: One of the sets of Nostratic phonemes that Bomhard has reconstructed is *tl' and *tl^[h]. One could well ask why Bomhard has not reconstructed *dl to correspond to the voiced series of stops in his reconstruction.

However, as we shall see, it hardly matters. The reconstruction of *tl' and *tl^[h] is based vaguely on PAA (or better, Semitic) phonological processes; and is unnecessary for Nostratic.

Bomhard shows as correspondences for initial Nostratic *tl': Proto-IE *k'; Proto-Kartvelian (no entry); Proto-Afroasiatic *tl'; Proto-Uralic *D^y; Proto-Dravidian *t; Proto-Altaic *k; and Sumerian d; and for initial Nostratic *tl^[h]: Proto-IE *k^[h]; Proto-Kartvelian *x; Proto-Afroasiatic *tl^[h]; Proto-Uralic *D; Proto-Dravidian *c; Proto-Altaic *k; and Sumerian (no entry).

It is very coy of Bomhard to predict AA correspondences identical to the Nostratic antecedents but, we can cut through the subterfuge and identify Arabic as a correspondent for Nostratic *tl' and Arabic as a correspondent for Nostratic *tl^[h].

In order to set this all aright, we will show that Nostratic *tl' should be amended to *k?; and Nostratic *tl^[h] should be amended to *k(h); in addition, we will show that the correspondences for these amended values in the daughter languages are exactly those we have predicted for them elsewhere in this critique.

(205)PN *tl^[h]ir^y-/*tl^[h]er^y- "to grow, to grow up, to thrive, to flourish" > PIE *k^[h]er-/*k^[h]or-/*k^[h]R- "to grow, to grow up, to thrive, to flourish"; PAA *tl^[h]ar-/ *tl^[h]6r- "to grow, to mature"; PD *cer..- "to thrive, to flourish, to grow, to grow well, to prosper, to be fertile, to increase, to be superabundant"

Not only Bomhard but also Pokorny has misinterpreted the semantic core of this root. The main idea here is revelation; and 'nourishment' is what is revealed with outer coatings are stripped off. Rather than 'grow' in the sense of 'become larger', the idea is to 'reveal an inner potential'. And, therefore, IE (2.) *k^er-, 'grow, make grow, nourish', is really the same root as (4.) *k^er-, 'damage', since the basal idea here is also that the outer covering falls away. Greek keraízo, 'devastate, plunder', shows us the essential idea of 'stripping away'.

In another publication, Bomhard specifies as a derivation Arabic šarakha, which my dictionary defines as 'be in the bloom of life', and Bomhard (re-)defines as "to become a youth, to mature, to grow up"; is this a definition à la Bomhard? The correct cognate with this šara- is seen in Egyptian š3', 'begin', and in IE *g^weru- in Greek brúo:, 'put forth shoots'. The kernel idea here is 'initiation' not 'growth'.

I can find no mention of Dravidian "*cer..- "to thrive . . ."" at Starostin's website. But, in any case, better cognates are available.

These cognates can therefore be easily related under one Nostratic form: *khxer^w-, and be reconstructed as: PL *KHXE-*RE-(*?A-), 'shed-apply'(+stative) = '(be) husk(ed), pick(ed) off, reveal(ed), cause(d) to appear'.

Presently, no corresponding Altaic cognate can be identified.

Accordingly, we amend:

(205)PN *khxer^y(re?-) "to husk(ed), to pick off (to be picked off), to reveal (to be revealed), to cause to appear (to be caused to appear)" > PIE (2.) *k^er-/*k^or-/*k^R- (for **k^her-/**k^hor-/**k^hR-) "to grow, to make grow, to nourish", *k^re:- (for **k^hre:-) "to be picked"; (4.) *k^er-/*k^or-/*k^R- (for **k^her-/**k^hor-/**k^hR-) "to damage, to fall apart", *k^re:- (for **k^hre:-) "to be crumbling (on the surface)"; PAA *ghar "to pick, to husk, to reveal"; Arabic in gharrara "to come forth"; in maghriz-un "grove"; in gharasa "to plant"; in Egyptian H3.t "food"; in H3.w "vintage"; in s-H3 "to strip, to reveal"; PD *kir.(i)- "to pluck"; *ki:r- (khxere?-) "herbs, greens"; possibly in *kir- "small, young (if = "naked")"; PU *kere-² "bark"; S kir₃ "to nip off"; ^(G[~]IŠ)kiri₆ (khxere?-) "orchard, garden" (source: PL *KHXE-RE[-?A])

An interesting example of the close relationships among the phonemes can be seen by comparing the forms and meanings under root #223 with those of this root.

DISCUSSION: A Nostratic root that differs from the previous one only by the stem-vowel is found as Bomhard's root #209 though Bomhard does not have the correct one. However, because of his methodological mistakes in reconstructing Nostratic phonology, he has it again as #246.

(209)PN *tl^[h]ar^y-/ *tl^[h]6r^y- "to cut, to cut into" > PAA *tl^[h]ar-/ *tl^[h]6r- "to cut, to slice"; PA *kar^y- "to scratch, to dig"

and

(246)PN *k^[h]ar-/ *k^[h]6r- "to cut" > PAA *k^[h]ar-/ *k^[h]6r- "to cut"; PA *ker-ti- "to cut into, to carve, to notch"

The semantics of this root are refreshingly straightforward. The sharp edges of shells were used as early cutting tools; and this root preserves the description of this archaic method of cutting.

The Arabic cognates for the first root, Bomhard, in another publication, mentions as šaraHa, "to cut in slices", and šarama, "to split, to slit". These words are related to IE *k^wer-, 'make, shape', seen also in Sumerian hur, 'scratch, draw, sketch, inscribe, outline, grind'. The Nostratic root is *xhor^y- (PL *XHO-*RE-, 'shallow-scratch' = 'lightly scratch').

For the second root, Bomhard, in another publication, mentions no Arabic cognates but instead cites a Proto-Semitic *k^[h]ar-at^[h]-, "to cut off, to cut down", based on Akkadian karâtu, "to hew off", among a few others. These words are related to IE **ger-, 'nibble off, notch', seen also in Egyptian k3w 'food'. The Nostratic root is *k?ar^y- (PL *K?A-*RE-, 'jaw-apply' = 'nibble off'). The Akkadian word is mirrored in IE *gred-, 'scratch' (better 'nibble off'; cf. Albanian gërrëse 'rasp'); and is seen again in *gras, 'eat (animal), nibble'.

Bomhard has a root (#216: *tl'ar-as-/*tl'6r-as-, "to bite, to gnaw"), to which he ascribes, in another publication, IE *gras- mentioned above. The Arabic "cognate" he offers there is Darasa, which means 'bite down strongly', which is probably related to IE *deres-, 'score'..

The correct cognates can therefore be easily related under one Nostratic form: *khxer^w-, and be reconstructed as: PL *KHXO-*RE-, 'mollusc(-shell)-apply' = 'cut off'.

Accordingly, we amend:

(209 & 246)PN *khxor^y "to cut off" > PIE (2.) *ker-/*kor-/*kR- (for **kher-/**khor-/**khR-) "to cut"; PAA *ghar- "to cut"; Arabic in gharrara "to cut"; in Egyptian x3j "to measure (if = "to cut off" or "cut marks on")"; Egyptian x3 "office (= "section")"; in Egyptian x3.yt "slaughter, massacre"; PD **kur- "to cut"; in *kUr-ad.- "trunk, log, stump" and *kurc- "stump"; PA *ku:r- "to cut out; sharp"; PU *kur3-⁴ "knife"; S kur₅ "to cut off" (source: PL *KHXO-RE)

It should be remarked that Bomhard does not offer a single Sumerian cognate for the thirteen Nostratic roots beginning with his *tl^[h]; that, if no other reason, should be enough to condemn this reconstruction as misbegotten.

DISCUSSION: Equally poorly formulated is Bomhard's proposed Sumerian correspondence to his Nostratic phoneme *tl', which, he claims is d. We will see that it is not.

(215)PN *tl'im-/ *tl'em- "to join, bind, or unite together" > PIE *k'em-/*k'om-/*k'M- "to join together, to unite"; PAA *tl'am-/*tl'6m- "to join together"; PU *D^yimä "glue"; S dim "band, binding; rope, cord, knot"; dim-ma "to tie together, to fasten, to bind"; dim-ma₂ "band, rope, cord".

and

(133)PN *t'im-/*t'em- "to make, to fashion, to create, to build" > PIE *t'em-/*t'om- "to build, to construct"; *tom-o-, *tom-u- "house"; S dim₂ "to make, to fashion, to create, to build"

Let us look at the second root first. As readers will know by now, I agree with the reconstruction of a Nostratic phoneme which has the characteristics of an apical glottalized voiceless stop, which Bomhard indicates as *t', and I indicate as *t?. Though Bomhard uselessly transcribes the IE equivalent as *t' also, his *t' always corresponds to the standard IE *d; and that is how I will transcribe it.

The IE root which seems to be involved here is Pokorny's *dem(6)-, 'construct, fit together'. An alleged extension of this root is found in IE *dem6-, 'tame, control, young animal', which Pokorny interprets as 'tie to the house, domesticate'. Nostratic *t?a/e appears in Egyptian as d, *t?o as t. There are therefore two Egyptian words of interest: dmj, 'join', and tm, 'be complete, close (mouth)'. Since tm represents Nostratic *t?om- or *thom-, Arabic Damma (*d.amma), 'draw together, collect, get hold of, embrace', is of interest since Arabic D is the predicted outcome of Nostratic t?o. Amazingly, Bomhard correctly identifies this Arabic cognate in another publication.

Sumerian dumu is 'child'; and first let us notice that retention of a final vowel in a CVC root is not usual. This indicates the former presence of a former 'laryngeal' which has lengthened the final vowel before disappearing (**dumû). This is, of course, what the *6 in *dem6-/*dem(6)- represents also; and it further suggests a connection. I cannot accept the idea that a culture regarded a 'child' as 'one constructed' or 'one fit together' but very well can accept the idea of 'one controlled', in effect, 'minor'; or 'embraced one'. This idea suggests the stative, and a Nostratic *t?omo?- would account for Sumerian **dumû.

The sign which reads tumu₂, 'bring, be suitable', also reads du; and, although an alternate reading of **dumu₂ is not recorded, the sign which reads tum, and which means the same, does have an additional reading of dum. I therefore amend tumu₂ to **dumu₂ and tum in this meaning to **dum.

We also have Sumerian dum₃, written with the signs for 'male' and 'female', and supposedly meaning either 'wife' or 'husband'; obviously, this means 'spouse', and to interpret it as 'one brought together/united with' seems very appropriate.

However, we must also deal with the root that underlies Sumerian dim and ^(G[~]IŠ)dim₂. The second Sumerian word means 'make, build, fashion, create'; and from the determinative, which means 'branch', it is clearly that 'construction' by means of 'tying' is contemplated. With this understanding, we can look at dim, which means 'make fast', and understand that this is really the same word applied to actual building 'construction' and other applications where something will be 'made fast by wrapping'. Now the distinction Bomhard seeks to make above (dim "band, binding; rope, cord, knot"; dim-ma "to tie together, to fasten, to bind"; dim-ma₂ "band, rope, cord") is ridiculous to anyone who has the slightest knowledge of Sumerian. Bomhard implies here that dim should only be understood as a nominal when it verbal use ('make fast') is clearly established. What is clear is that dim-ma should be interpreted as a mater lexionis for an unrecorded reading of **dimmâ, a stative form of dim, which means 'tied together, fastened, bound'. With this insight, we can relate dim and **dimmâ and dim₂ to IE *dem- and *dem6-, 'construct' and further with Egyptian dmj, 'join'.

We can therefore see that Uralic "*D^yimä "glue"" means not 'glue' per se but rather 'that with which something is made fast' (we will see below that the Nostratic root for 'glue' is *nkem- in Egyptian qmj, 'gummed'; qmy.t, 'gum resin'; qm3, 'create' [**'cause to adhere together']). Uralic *D^yi (D is for 'delta') is therefore to be regarded as a strictly Uralic reaction to a Nostratic *t?e: palatalization and fricativation, not requiring a separate Nostratic phoneme to which to correspond.

These cognates can therefore be easily related under two Nostratic forms: PL *T?O-*MO-(*?A-), 'put-together-slide'(+stative) = 'bring together'('brought together').

and: PL *T?E-*MA-(*?A-), 'spin-around-stay' = 'make fast by wrapping'('made fast by wrapping').

Accordingly, we amend:

(215)PN *t?om^w-(/mo?-) "to bring together ("to be brought together"), to control ("to be controlled")" > PIE *dem-/*dom-/*dM- (*dem6-/*dom6-/*dM6-) "(to bring together), to tame, to control"; PAA *Dam- "to bring together"; Arabic in Damma "to draw together, to collect, to get hold of, to embrace"; in Egyptian tm "to close (mouth), to complete"; S tum (for **dum) "to bring, to be suitable"; tumu₂ (for **dumû₂) "to bring, to be suitable" (source: PL *T?O-MO-[?A-])

and

(133)PN *t?em-(/ma?) "to make fast by wrapping ("made fast by wrapping")" > PIE *dem-/*dom-/*dM- (*dem6-/*dom6-/*dM6-) "to construct, to fit together"; PAA *tam(/ma?-) "to cut"; Arabic in tamma "to complete (and by 'completing', "fix"), "to executive (an order) [="to adhere to"]"; in Egyptian dmj "to join, to cleave to"; PU *D^yimä "glue ("that with which something is made fast")"; S dim "to make fast"; dim-ma (for **dimmâ) "made fast, tied together, fastened, bound, *wound"; ^(G[~]IŠ)dim₂ "to make, to build, to fashion, to create, *to construct" (source: PL *T?E-MA[?A-])

This, of course, leaves us with Bomhard's "PIE *k'em-/*k'om-/*k'M- "to join together, to unite"" to identify. As we have seen, an IE transcription of *k' is useless sense it will always appear in IE as *g, and so will we transcribe it. Equally amazingly, Bomhard has identified, in another publication, the correct IE cognate of the root we will propose to add: IE *g^em-, 'to marry', which we will more narrowly define as 'to cause to adhere'.

Above we mentioned Egyptian qmj, 'gummed'; qmy.t, 'gum resin'; qm3, 'create' (**'cause to adhere together'). To that, we can now add the Sumerian cognate: gim, 'make, do, create', the sign for which probably indicating a tool for cutting plants to induce bleeding of sap. How Bomhard could have missed this Sumerian word corresponding to the two above-mentioned roots is a matter of extreme puzzlement.

This set of cognates can now be easily related under PL *QE-*MA-(*?A-), 'congeal-stay'(+stative) = 'glue together'('glued together').

Accordingly, we amend to include:

(Addition AQ)PN *nkem-(/ma?-) "to glue together ("glued together")" > PIE *g^em-/*g^om-/*g^M- (*g^em6-/*g^om6-/*g^M6-) "to be married"("to be glued together"); PAA *qam- "to glue together"; Arabic in qama?a "to stay in (a place)"; Egyptian in qm.j "gummed"; qm.yt "gum, resin"; qm(-)3 "to create ("to cause to adhere together")"; PU possibly in *kämä- "firm"; S (n)g[~]³im "to make, to do, to create"; (source: PL *QE-MA-[?A-])

It should be noted that Bomhard, as mentioned above, asserts that Sumerian d is the response to Nostratic initial *tl'. There are four examples with this initial in his dictionary. Only the first (sic!), discussed above, has any Sumerian "equivalent"! No responsible comparativist would make an assertion of correspondence based on one example; and, further damningly, we have seen, that example is wrong. Frankly, this is absolutely indefensible; and I can readily understand why Bomhard would choose not to attempt to defend his 'work'. What I cannot understand is why Bomhard wrote it in the first place.

DISCUSSION: We will now look at the remaining two entries for which Bomhard reconstructs initial Nostratic *tl'.

(217)PN *tl'al-/*tl'6l- "to be bent, curved, round" > PIE *k'el-/*k'ol-/*k'L- "bent, curved, round"; PAA *tl'al-/*tl'6l- "to be bent, curved, round"

and

(218)PN *tl'uk^[h]-/ *tl'ok^[h]- "to push, to shove, to thrust in" > PFU *D^yukk3- (*D^yokk3-) "to put (in), to stick, to thrust (in)"; PD *tuk- "to push, to shove"

The first root is connected with (1.) *gel-, 'form into a ball, ball up, what is rounded, what is spherical'. Since it is not palatalized, the Nostratic vowel should be *a or *o. Without finding a cognate in a language that gives us a further indication of the Nostratic vowel-quality, it is not possible to specify it. In another publication, Bomhard mentions no cognates that would enable him to specify Nostratic *a so his reconstruction has no more value than a random guess. However, coincidentally, his guess is correct (though he had no warrant to make one). We connect Egyptian qnj, 'embrace, bosom'. Because Egyptian q can only represent Nostratic *nka/e or *nkha/e, the Nostratic first syllable is *nka since *nk(h)e is ruled out by the lack of palatalization in the IE correspondent, which, also, being voiced, eliminates *nkha. This word is possibly found in Starostin's Altaic *giá:l'u-, 'ring, bracelet'.

The Arabic "cognate" Bomhard cites in another publication is Dali¿a, which my dictionary translates as 'be crooked'. In fact, the basis for it, Dalla, 'stray, deviate, be misled'. Thus, the main idea is not 'bending' as such but rather 'deviation from an expected trajectory'. The corresponding IE cognate is (2.) *del-, 'wobble, sway'; and we see it again in Egyptian tnbX, 'turn aside, swerve'; and tnm, 'turn aside, go astray, be confused'. Egyptian tnmw, 'beer', is 'what makes one sway'. And finally, tnj, 'grow old' ('begin to wobble'). Very speculatively, a Sumerian sign with many archaic antecedents and meanings, has the meaning 'waver', and the possible reading dul₅; this might possibly also be related. The Nostratic root here is *t?ol-, 'sway' (PL *T?O-*NHA-, 'torso-move back and forth' = 'sway').

These cognates can therefore be easily related under two Nostratic forms: PL *QA-*NHA-, 'semi-circular-move back and forth' = 'make round'.

Accordingly, we amend:

(217)PN *nkal- "to make round" > PIE (1.) *gel-/*gol-/*gL- "to form into a ball, to ball up, what is rounded, what is spherical"; PAA *qal- "to make round"; Arabic in qalt-un "eye-cavity"; in qalada "to put a necklace on"; in qilf-un "tree-bark"; in Egyptian qnj "to embrace; bosom"; PA possibly in *giá:l'u- "ring, bracelet"; S gala₂ (for **(n)g[~]³ala₂; *nkala?, "made round, rounded") "storage-pit" (source: PL *QA-NHA-)

If Bomhard's correspondences had any validity, then we should expect an IE equivalent for root #218 beginning with *k', i.e. conventionally represented IE *g. However, we have seen that such a correspondence is pure fantasy. And, as we have also far too often seen, it appears that Bomhard has conflated two separate roots of somewhat related meanings.

Firstly, let us look at "PFU *D^yukk3- (*D^yokk3-) "to put (in), to stick, to thrust (in)"", which Rédei has as *D^y8kk3 (with ^y for '). For comparison, we should first look at IE *dhe:ig^w-, 'stick into, stick in, set firmly'. Based on comparison with *(s)teig-, 'stick into', which I suspect is an s-mobile form of the same root, I question the final *g^w of *dhe:ig^w-. Pokorny gives examples of this word in Latin (fi:go:), Lithuanian (and Latvian), which have g and Germanic languages which have *k. Thus, its reflexes fit the profile of IE *g as well or better than IE *g^w. I am aware of Archaic Latin fi:vo:, the datum on which this reconstruction of *g^w apparently is based, and would counterpose that there is no trace of the velar glide in the derivates of the other languages. The explanation for it that most readily occurs to me is that an adjective in *-u (cf. Lithuanian dygùs, 'pointed') has contaminated the simplex.

In other studies, I have attempted to demonstrate that IE final *bh, *dh, and *gh, derived only from Nostratic *p?fa, *t?sa, and *k?xa (and therefore, without glide in the Pontic stage of development), frequently lose their aspiration, becoming *b, *d, and *g.

Based on what I perceive to be the Egyptian cognate, 'H3 (for **'jH-3), 'fight ('use a spear or arrow')', since H goes back to Nostratic *k?/ha/e, I amend *dhe:ig^w- to **dheig(h)- (the long vowel being a compensation for loss of the final aspiration) and *(s)teig- to *(s)t(h)eig(h)-.

IE *dhe:ig^w- also show what is considered to be an Ablaut variation: *dho:ig^w-. I believe this is better to be explained by supposing two Nostratic roots of similar meaning, one with an *e and another with *o vocalisms, which have been conflated by Pokorny.

We are now in a position to provisionally reconstruct two Nostratic roots: *t?seyek?x- and *t?soyek?x-.

For *t?seyek?x-, the correspondences advanced in this critique by me would predict a Sumerian equivalent of **zîk; and, in fact, the Sumerian sign which means 'set firm' has the reading zig/k though it is usually read zid (cf Old Indian sthíti-, 'standing') in this meaning.

Accordingly, we reconstruct: PL *T?SE-*¿E-(*K?XA-), 'finger-like=(arrow-)bolt+hang' = 'be sticking in'.

And we amend:

(218)PN *t?seyek?x- "to be sticking in, to be firmly set" > PIE *dhe:ig^w-/*dhi:g^w- [for **dheig(h)-/**dhig(h)-] "to be sticking in, to be set firmly"; S zig (for **zîk[?]) "to be firmly set"; (source: PL *T?SE-¿E-K?XA)

But, of course, that leaves Nostratic *t?soyek?x- for which to account.

As we have mentioned above, we predict Nostratic *t?/ho and *t?/hso will appear in Sumerian as tu. For this root, we would theoretically have **tuik but I believe we can see that this became first **t^wik then *tîg/k. The Sumerian sign which reads tig/k, in combination with ge₁₆, i.e. tig-ge₁₆, is glossed as Akkadian muntahzu, 'battler'; and this corresponds roughly to Egyptian amended **'jH(-)3, 'fight'. In addition, a sign which means, among other meanings, 'strike dead', when duplicated, reads Sumerian dig₃, which we would amend to **tîg_x

Accordingly, we reconstruct: PL *T?SO-*¿E-(*K?XA-), 'branch-like=(spear-)shaft+hang' = 'stick into'.

and we amend to include:

(Addition AR)PN *t?soyek?x- "to stick into" > PIE (2.) *dho:ig^w-/*dhi:g^w- [for **dhoig(h)-/**dhig(h)-] "to stick into"; PAA *dayaj "to stick into"; in Egyptian 'H(-)3 "to fight"; PFU *D^y8kk3- "to put (in), to stick, to thrust (in)"; S in tig-ge₁₆ (for **tîk) "battler"; dig₃ (for **tîk_x) "to strike dead" (source: PL *T?SO-¿E-K?XA)

It will be noticed that we have still not accounted for "PD *tuk- "to push, to shove"". I cannot find sufficient cognates to be fairly confident in a reconstruction of this root however there is IE *(s)teug-, 'push', which is very probably related. The only point to be drawn from this is that Bomhard's equation of Nostratic *tl' = IE *k' is almost certainly wrong; and with it, his correspondence of Nostratic *tl^[h] = IE *k^[h].

DISCUSSION: Having disposed of a few of Bomhard's phantom Nostratic phonemes, we now turn to the 'laryngeals', or better the laryngals and pharyngals, which, to his credit, he correctly identifies as *?, *h, and *¿, H (the last, my web-convention for dotted h or top-barred h).

These roots have their own challenges though, if we correct identify an Arabic cognate, we can precisely identify the Nostratic phoneme.

There are 57 roots in Bomhard's dictionary beginning with a laryngal or pharyngal. Of those, Bomhard offers only seven Sumerian equivalents. This is partially because Bomhard incorrectly expects Sumerian h from Nostratic *H, which, according to my series of correspondences, shows up as Ø as do all the others. In addition, Bomhard has invented out of whole cloth an IE phoneme which he notates as *Hh. This 'phoneme' can safely be ignored as just another aberration since, like the others, it has no effect on the derived IE forms, with the exception of *?, but to lengthen and so preserve the original Nostratic vowel quality.

(385)PN *Ham-/*H6m- "to be sharp, sour, acid" > PIE *Hhem- [*Hham-] / *Hhom- "sharp, sour, acid"; PAA *Ham-/*H6m- "to be sharp, sour, acid"

This root probably has a Sumerian correspondent in Sumerian um₂, which is written with the sign for 'sun'. However, it must be admitted that this meaning is not documented for it.

These cognates can therefore be easily related under PL *HHO-*MO-, 'hot-overall' = 'very hot'.

Accordingly, we amend:

(385)PN *hhom^w- "to be very hot" > PIE (1.) *om-/*6m- (for **o:m-; cf. *o:mo-s) "raw, bitter"; PAA *hham- "to be very hot"; Arabic in hamma "to heat water"; in Egyptian hm "to be burning"; PA *umV- "fire; to burn"; S um₂ (for **ûm₂) "to be hot[?]" (source: PL *HHO-MO)

DISCUSSION: Bomhard normally has Arabic H as the correspondent to Nostratic *hh (his PN *H) but in the case of the root below, without any explanation or discussion, he, in another publication, connects it with Arabic (xaqq-un, 'crevice in the ground')

(401)PN *Hak'-/*H6k'- "to cut into" > PIE *Hhek'-w(e)siH- [*Hhak'-w(e)siH-] "ax"; PAA *Hak'-/*H6k'- "to cut into"

This root a Sumerian correspondent in Sumerian ag, 'cultivate', the archaic sign for which pictures a 'runnel with water'.

This is, of course, the same root as Bomhard's #396:

(396)PN *Hak'-/*H6k'- "field" > PIE *Hhek'-ro- [*Hhak'-ro-] "'field"; PAA *Hak'-/*H6k'- "field"

And it will be treated under the same amendment. Sumerian has agar_(2,3), 'field; commons', which suggests strongly that IE *ag^-ro-, 'field', may have been incorrectly reconstructed by Pokorny. In view of girim, 'piece of clay; detach a piece of clay', and im(i), 'clay', it seems reasonable to abstract **gir_x, 'split', even though it is not attested in this meaning. Of course, this implies that there is a Nostratic root corresponding to **gir_x, and we will investigate it in the next section.

These cognates can therefore be easily related under PL *HHA-*K?E-(*RE), 'water-split(-apply)' = '(make) a runnel (for irrigation)(,field)'.

Accordingly, we amend:

(396 & 401)PN *hhak?^y-(/k?er^y) "(to make) a runnel (for irrigation)(; irrigated field)" > PIE in *ag^-ro- (for **ag^er-) "field"; PAA *Haj- "to make a runnel"; Arabic in hajja "to probe a wound"; in Hajama "to scarify"; in Egyptian jk.w "stone-quarry"; jk.y "quarryman"; PA *ák'a- "crack, opening"; in *àk'u- "to dig, to delve"; PU possibly in *äk(-)t3- "to beat, to hew, to cut"; PD agar.- "to dig"; S ag "to cultivate"; agar_(2,3) (for **agir_(2,3)[?]) "field; commons" (source: PL *HHA-K?E-[RE])

DISCUSSION: The Nostratic root corresponding to Sumerian **gir_x is, though presented with an incorrect stem-vowel, seen in Bomhard's:

(282)PN *k'ar-/*k'6r- "to cut: to cut into, to make an incision, to engrave, to notch; to cut off, to sever, to nip off, to clip; to cut in two, to split, to bite" > PIE *k'er- /*k'or-/*k'R- "to cut; to cut into, to make an incision, to engrave, to notch, to cut off"; PK *k'Rt'-wN- "to peck, to bite"; PAA *k'ar-/*k'6r- "to cut; to cut into, to make an incision, to engrave, to notch, to cut off, to sever, to nip off, to clip; to cut in two, to split, to bite"

In another publication, Bomhard connects this root with *gerebh-, 'scratch'; and I believe he is correct. Also, this word is quite probably cognate with Arabic karaba, 'plough', though Bomhard incorrectly assigns Arabic qar- to this root. These Arabic words are possibly masked in IE (2.) *ker-, 'cut' (Nostratic *nkhor^y-, 'pinch [off]'); and Sumerian (n)g[~]³ur₅, 'circumcise (?); trim away, strip; cut, clip (a part of the body); notch, incise; fell trees; be parted, relieved of').

These cognates can therefore be easily related under PL *K?E-*RE, 'split-apply' = 'split'.

Accordingly, we amend:

(282)PN *k?er^y- "to split" > PIE in *g^er- /*g^or-/*g^R- "crumbling, to become ripe, to age"; in *gerebh (for **g^erebh-) "to scratch"; PAA *kar- "to split"; Arabic in karaba "to plough"; in karada "to cut"; in karâ (k-r-y) "to dig (a canal)"; in Egyptian s(-)k3 "to cultivate, to plough"; S **gir_x "to split" (source: PL *K?E-RE-)

DISCUSSION: A Nostratic root for which Bomhard has reconstructed the incorrect 'laryngeal' is:

(370)PN *¿uw-/*¿ow- "flock or herd of small animals; sheep and goats" > PIE *Hhowi- "sheep"; Afroasiatic: Egyptian 'w.t "sheep and goats, animals, flocks, herds"; PFU *u-t^ye (from *uwi-t^y[e]) "sheep"; S u₈ "ewe"

First, we need to investigate Bomhard's proposed cognate Egyptian 'w.t. Antonio Loprieno's Ancient Egyptian — A linguistic introduction was written and published in 1995, a year before Bomhard's dictionary. In it, on page 31, Loprieno equates Egyptian 'r.t, 'portal', with Semitic *d-l-t-, and Egyptian 'ff, 'fly', with Semitic *dh-b-b. Most scholars believe that Egyptian ' later developed a pronunciation of [¿] in Egyptian ( I believe the evidence favors a pronunciation of [?]) but however it came to be pronounced, for comparative purposes, /'/ was an apical not a pharyngal obstruent. Thus, Bomhard, operating with out-of-date phonology, wishes to compare Egyptian words in ' with AA (Semitic) words in ¿; and this is absolutely wrong unless we are dealing with a loanword borrowed into Egyptian after the earliest pronunciation of ' was lost.

It will be noticed that Loprieno has both Semitic *d and *dh for Egyptian '. I have written above that I believe Egyptian ' represented Nostratic *t?so and *thso, for which I believe the Semitic equivalents were Z and S. What is oddly coincidental is that I believe Nostratic *t?sa/e and *thsa/e would result in Semitic d and dh as I have outlined above. Obviously, there is a problem here. I find myself agreeing with Loprieno in principle but disagreeing with the detail. In my opinion, Semitic *d-l-t- should be compared with IE *dhel- (Nostratic *t?sol-), 'arch(ed passage)' (Sumerian tul, '**cover, shade'; another reading of the sign presently read dul). There are Arabic roots in Zal- that can be understood as meaning 'cover' (e.g. Zallala, 'shade, give shade over)'. Egyptian 'r(r.w).t, 'portal' (Sumerian tul₂, 'well, canal, cellar, worm'), on the other hand, should be compared with IE **tor- (for **thor, Nostratic *thsor^w-), 'hole', in *tormo-s, 'hole', from (3.) *ter-, 'bore through'. There are Arabic roots in Sar- that could be understood as meaning 'go/bore through' (e.g. Sarada, 'hit the target [with an arrow])'.

The circumstances surrounding Egyptian 'ff, 'fly', and Semitic *dh-b-b are rather involved. Rather than a biliteral root with a geminated final, I believe this word should be analyzed as a compound of Nostratic *p?op?^w, 'all-swollen'='insect', found in IE *beb- (cf. Swedish bobba, 'insect'); preceded by the Semitic reflex of Nostratic *thsa-, 'ass', i.e. *dh; yielding 'equid' or 'horse-fly', also called 'botfly' (Nostratic *thsap?op?^w-). The Egyptian form, 'ff, seems rather to be a compound with a related Nostratic *thsawa:-, 'asses', which has been resolved to *thso(:)- (*thso[:]p?op?^w-), so that the two cannot be directly compared.

We may now consider the what cognates "'w.t "sheep and goats, animals, flocks, herds"" might actually have. It is, I think, obvious, that the basic meaning of this word pertains to 'large collections' rather than to any specific animal. This is ultimately a derivative of Nostratic *thso, 'swarm', and here we have its plural: *thsowa:-. Traces of this root are few in Semitic but we can point to Arabic Siwâr-un, 'herd of oxen'. The IE cognate is much better represented: *teuta:- (for **theuta:-), 'people'.

The most convenient way to approach an analysis of this word is through its Proto-Language components. PL *SHO designated a small group of humans moving together as a group: a 'clan'. It was also applied to animals that traveled in similarly small groups like sheep and goats; in this application, it meant simply '(small) herd'.

With an extension, PL *SHO-*FHA, 'follow-move-repeatedly' = 'following' = 'clan', provides the basis for IE *swe- in words for 'clan'; with *P?FE, formant of place names, it produces 'clan-place' = '(place of) assembly' (IE *swebh-).

In its simplest form, we can see *SHO in Egyptian s, 'sheep', which, of course, really just mean 'herd (animal) par excellence'.

Therefore, it is to be anticipated that IE *owi-s-, 'sheep', was originally an *s-stem, and that the *owi- is a qualification that narrows the definition from 'herd-animal' to specifically 'sheep': i.e. **owis-. That it was an *s-stem is supported by Sumerian us₅, 'mother ewe, adult female sheep'. It is also supported indirectly by the IE feminine form, *owika:-, which, in my opinion, has the same first qualifying element, followed by -ka:-. The final -a: of this element is due to PL *HA, 'female'; the first element is *KHA, 'rutting male', which has been transferred to 'goat/sheep'. It is also seen in IE *kag^(-)o- , 'goat', when combined with *K?E, 'penis' = 'male', preceded by *HA, stative, but possibly 'female' (*ka:g^- shortened to *kag^- + *o-).

The Nostratic sequence *w(Vy) becomes g[~]V in Sumerian similar to what we find in Armenian: IE *woinyom- = Armenian gini, 'wine'. Therefore the Nostratic parent form cannot have had this sequence. However, Nostratic *hhowa: ( PL *HHO-*FHA), '(mal)odorous', would appear in Sumerian as u (really **u:); and, with an adjective formant, -*y ( PL *¿E-), this would produce IE *o:wi/i:-. The strong smell of sheep and goats is proverbial.

From Sumerian (us₅), it appears that one Nostratic word for 'sheep' was *hhowa:s^w-, '(mal)odorous (small) herd', which was expanded for IE to *hhowa:ys^w-; and there are faint traces in IE of words related to 'sheep' that seem to reflect the earlier (or alternate) form without -*y: e.g. Greek óa, 'fleece', vs. oía; and OHG ou, 'sheep', vs. ouwi; Latvian àuns, 'ram', vs. avins.

However, it is also clear that Nostratic *hhowa:- by itself was a designation for 'sheep' as Sumerian u₈ shows. Sumerian ua, another reading of the sign for u₈, is either a better rendition of *hhowa:- or represents an addition of *-HA, feminine formant, to the root (*hhowa:h-). In addition, there is udu (for **utu; cf. utua₂, 'ram', and utul₃, 'shepherd') , 'sheep; small cattle; ram, wether', corresponding roughly to AG e:owde, 'sheep-herd'.

Just for the sake of completeness, let me mention that the oldest name for 'sheep' was apparently simply *PHFE-, again identifying the primary characteristic of 'sheep' as 'bad odor'.

These cognates for IE *owi-s-, 'sheep', can therefore be easily related under PL *HHO-*FHA(-*¿E-)(*SHO), 'odor-move-repeatedly='(mal)odorous'(-'like'='(mal)odorous')('[small-]herd') = 'sheep'.

Accordingly, we amend:

(370)PN *hhowa:(y)(s^w-) "sheep" > PIE in *ówi(s)- "sheep"; first element in *owi(-)ka: "ewe"; PAA **(hhawaya)sa- "sheep"; second element in Egyptian s "sheep"; PFU *u(-)ce- "sheep"; S first element in u₈ (for **û₈) "ewe"; us₅ (for **ûs₅) "mother ewe, adult female sheep" (source: PL *HHO-FHA-[?E-][SHO-])

Nostratic *hho-, 'odor', is found as the first component of IE *od- (from **o:d-; PL *HHO-*T?A, 'odor-give' = 'smell'), 'smell'.

DISCUSSION: Contrasting with Nostratic *nkher-, '(ovine) horn', is the following root:

(437)PN *?ar-/*?6r- used as the base for the designation of various animals > PIE *?er-/*?or-/*?R- used as the base for the designation of various domestic horned animals; PK *arc[^h]k[^h]w- used as the base for the designation of various animals; PAA used as the base for the designation of various animals; PD *er- "bull, bullock, ox, buffalo"

In my opinion, Old Irish heirp, 'fallow-deer', points to the original significance of this root. Included by Pokorny under this heading is Middle Irish fearb, 'red deer', which one may possibly reconstruct as a derivation of PL *FA-*RA, 'palm-tree='palmate antler, elk' (cf. IE *w_eren-, 'ram, sheep, lamb'; PAA *war-, 'bull, cow').

These cognates can be related under PL *?E-*RA, 'tooth-tree='(antler) rack'.

Accordingly, we amend:

(437)PN *?er- "(antler) rack, antlered animal" > PIE (2.) *er-/*em>or-/*R "goat, sheep, cow, sheep, fallow-doe"; PAA *?ar "ram, goat"; Egyptian, possibly in j3.wt "herds"; in *?arkh- "cattle" (cf. OHG irah "male goat"; PL *?E-RA-KHXE[?]); PD **ir- in Tamil iralai "stag, a kind of deer"; Telugu iri "stag"; and irri "antelope" (source: PL *?E-RA)

DISCUSSION: Another important root built on PL *?E, 'tooth', is the following:

(452)PN *?il-/*?el- "hoofed, cud-chewing animal" > PIE *?el-/*?ol- "hoofed, cud-chewing animal"; Afroasiatic: PSC (Proto-Southern-Cushitic) *?aale "hoofed, cud-chewing animal"; PD *il- "stag, antelope, deer" (Tamil iralai [from *ilar-] "stag, a kind of deer"; Telugu iri "stag"; irri [from *ilri] "antelope", le:t.i, le:d.i [from *ilat.i] "antelope"; Malto ilari "the mouse deer"); Altaic: Mongolian ili "a young deer, fawn"; Khalkha il "a young deer, fawn"

In my opinion, Bomhard is on the right track here but simply has not drawn the next appropriate conclusion, which is that the root fundamentally means 'chew the cud'.

Supporting this conclusion is IE (7.) *el- , 'be hungry', which I believe vividly portrays the involuntary grinding of teeth that accompanies extreme hunger. There is also IE *elg-, 'pitiful, needy', which corresponds to my reconstructed Afrasian **?alak-, the basis for Orel-Stobova's *?alVk-, 'bite, chew', and *?ilik-, 'tooth'. As we have seen, Nostratic *k? becomes Afrasian *k and IE *g, and this points the way to IE *elg-, 'pitiful, needy', which, on the basis of OHG ilki, 'hunger', I would suggest has a basic meaning of 'chew, grind the teeth in hunger'. The final element is probably PL K?A, which itself means 'jaw, chew'; I presume the combination began by meaning 'cud-chew', i.e. 'chew a cud'.

Starostin's website has no trace that I am able to find of Bomhard's proposed Dravidian *il-, 'stag, antelope, deer'; and I am unenthusiastic about Bomhard's proposed metathesis. To me, most of the forms he cites can just as easily to related to Nostratic *?er- discussed above as amended root #437; and I will place them there. I also reject le:t.i, le:d.i, 'antelope', as having little justification beyond superficial similarity; however Malto ilari, 'mouse deer', may possibly be cognate.

Similarly, it is probably unjustified to reconstruct an Altaic root based solely on Mongolian ili, 'young deer, fawn', and Khalkha il, 'young deer, fawn', but I will include it as a possibility since Starostin has Altaic *elV(-k'V)-, 'deer', at his website.

These cognates can be related under PL *?E-*NHA, 'tooth-move-back-and-forth'='chew (cud)'.

Accordingly, we amend:

(452)PN *?el- "to chew (cud), cud-chewing animal" > PIE (7.) *el-/*em>ol-/*L- (also *elk-/*em>olk-/*Lk-) "to be hungry"; (5.) *el-/*em>ol-/*L- "to destroy, to spoil (="to chew up")"; (1.) *el-/*em>ol-/*L- forms animal and tree names ("cud-chewer" and "what is chewed"); in *elk^- "deer and similar animals" (cf. OHG ëlho "elk"; PL *?E-NHA-KHXE); in *elg-/*olg-/*Lg- "pitiful, needy (="chewing"[?] = "hungry"; cf. OHG ilki "hunger"; PL *?E-NHA-K?A)"; PAA **?al- "to chew (cud)"; possibly in PSC (Proto-Southern-Cushitic) *?aale "hoofed, cud-chewing animal"; in **?alak- (Orel-Stobova *?alVk-) "to bite, to chew" (cf. PIE *elg- "pitiful, needy" {see above}); in **?alak- (Orel-Stobova *?ilik-) "tooth"; PU possibly in *äl3²- "tree-sap (if = "what is chewed; mastic")"; PA *elV(-k'V)- "deer"; possibly in *ele "fragrant grass (if = "what is chewed")"; probably in Mongolian ili "young deer, fawn", and Khalkha il "young deer, fawn"; PD **il- "stag, antelope, deer[?]" (possibly in Malto ilari "the mouse deer") (source: PL *?E-NHA)

DISCUSSION: Yet another important root built on PL *?E, 'tooth', is the following:

(418)PN *?at'-/*?6t'- "to chew, to bite, to eat, to consume" > PIE *?et'-/*?ot'- "to eat"; Afroasiatic: PS *?at-am- "to bite into"; Altaic: Mongolian ide "to eat, to feed on, to gnaw, to eat up, to devour, to consume"; Buriat ed'e- "to eat"; Dagur ide_v- "to eat"

What is truly strange here is that Bomhard, who cites other forms from Orel-Stobova, seems unaware of their root #83, *?et-, 'eat'. Similarly, Starostin's website has Altaic *ite, 'eat', which seems also to have escaped Bomhard's attention.

In another publication, Bomhard associates Arabic ?aTama, 'bite into', for which he provides no source, with this root but I am unable to find a reference to this alleged Arabic word. However, this word, if it exists, would have Nostratic *?Vt^w- as its basis, and so must be disassociated from this root.

This root, which, because of its very basic meaning, one might expect to be fairly well-distributed, cannot be found, at present, for Dravidian, Sumerian, or Uralic.

These cognates can be related under PL *?E-*T?E, 'tooth-grind'='eat'.

Accordingly, we amend:

(418)PN *?et?^y- "to eat" > PIE *ed-/*em>od- "to eat"; PAA **?at- (Orel-Stobova *?et-) "to eat"; PA *ite- "to eat" (source: PL *?E-T?E)

DISCUSSION: There are a number of important roots built on PL *HHA, 'water', of which one is the following:

(382)PN *hhaw-/*hh6w- "to sprinkle, to spray, to rain" > PIE *hhew-r- [*hhaw-r-]/*hhu-r-, *hhw-er-*hhw-or- "to sprinkle, to spray, to rain"; (n.) water, moisture", *hhw-ers-/*hhw-ors- "to rain"; *hhew-on(t^[h])- [*hhaw-on(t^[h])-], *hhew-N(t^[h])- [*hhaw-N(t^[h])-] "spring, well" (also used as the base of river names); Afroasiatic: Egyptian Hwj "to surge up, to overflow, to rain"; Hwyt "rain"; PD *var "flood, torrent, inundation"

Firstly, it might be interesting to notice that in 38 roots which Bomhard reconstructs with PN pharyngal *hh (#379-#416), which he notates as *h with a superior bar, fully eight examples have, under Afroasiatic, only an Egyptian "cognate" listed beginning with H (his dotted-h) or x (his hooked-h) — apparently this makes no difference to him. I think most readers would agree that this is an unusually high percentage (21%) of alleged AA roots which have preserved cognates only in this one branch of AA. Even more interesting is that of the remaining 30 roots, Bomhard provides only 13 with a proposed non-Egyptian AA cognate and an Egyptian one.

These facts taken by themselves should indicate that Bomhard is wrong to equate Nostratic *hh with Egyptian H or x though, of course, his proposal to link AA h. is correct. Here, unfortunately, consulting Loprieno would not have helped much since he also proposes that AA h. becomes Egyptian H. That this is incorrect is shown by the Egyptian cognates beginning with H and x listed in the Nostratic Dictionary - Alphabetical Listing under *K?X and *KHX.

Orel-Stobova does have this root (#1303) as "*h.Vw- "rain" (v.)" with non-Egyptian cognates as well as Egyptian Hwy, 'rain, flow', the latter which I reject as incorrect. The correct Egyptian cognate can be seen in jw(-)H, 'moisten, water' (PL HHA-FHA-K?XE).

The better interpretation of Egyptian Hw(j) is to regard it as a specialized application of 'beat', i.e.i.e. 'beat down on' as of rain. It is cognate with IE *ka:w-, '(be) beaten', which is a derivative of PL *KHXA-*FHA, 'hurt-do-repeatedly'='(be) beat(en)'. The preservation of the Nostratic medial A in IE is occasioned by the transient form **khaw-, the aspiration of which, when given up, lengthens and preserves the root-vowel. This initial element can also be seen in IE **kat- (from **khat-), 'fight', an alternate form of the root usually reconstructed as *k^at- (cf. Old Bulgarian kotora, 'war'), which is PL *KHXA-*THO, 'hurt-do-repeatedly'='(time of) hurting'.

An Arabic word which I believe is a cognate has been neglected by both Bomhard and Orel-Stobova: Hawiya, 'be or become dark green', which I believe is an adjectival formation, *HHA-FHA-*¿E, '-like' = 'become water-like in color'.

Dravidian alveolar *r in *var points to Nostratic *E, which makes probable the reconstruction PL *RHE, to be interpreted either as 'rain' or 'fall'.

This root cannot be found, at present, for Altaic or Uralic.

These cognates can be related under PL *HHA-*FHA, 'water-go-around'='moisten'.

Accordingly, we amend:

(382)PN *hhawa:- "to moisten" > PIE (9.) *aw(e)- "to sprinkle, to moisten, to flow"; PAA **Haw- (Orel-Stobova *h.VW-) "to moisten"; in Arabic Hawiya (*HHA-FHA-¿E) "to be or become dark green"; in Egyptian jw(-)H "to moisten"; PD in *var (*HHA-*FHA+*RHE "rain", possibly "to fall"; cf. PIE *awer- "[rain-]water ") "flood, torrent, inundation"; S in ur_2,3,4 (*HHA-FHA-RHE) "to flood" (source: PL *HHA-FHA)

DISCUSSION: One of the important usages for PL *HHA, 'water', is based on its characteristic of reflectivity; as a consequence, it is used in roots for 'bright' and 'shine', and by association: 'heat'.

(393)PN *hhaw-/*hh6w- "to shine" > PIE *hhew-s- [*hhaw-s-], *hhw-es-/*hhu-s- "to shine", *hhew-k'- [*hhaw-k'- "to shine"; PAA *Haw-/*H6w- "to shine"

Orel-Stobova does not have this root but Bomhard, in another publication, cites Arabic Hawira, 'be or become white', which I believe he has correctly identified as a cognate. On the other hand, he also cites there Egyptian Hwy, 'illumination, light', which Bomhard has found in Budge. This word is not listed in the Wörterbuch or Faulkner, and probably is not attested in Old or Middle Egyptian, which militates strongly against its use as a potential cognate here. Even if it does exist in Late Egyptian, it is probably to be referred to IE (2.) *k^ew-, 'kindle, burn up', rather than to this root.

One of the famous gods of Egypt is Osiris, the god of resurrected life, traditionally transliterated as wsjr(j). The plainest spelling of the name is a sign, generally described as a 'seat' (Gardiner's Q1), normally read st, which means 'seat, place'; and an 'eye', which is normally read jrj, and means 'make'. Unless it is a loanword, it appears to be a compound of X + 'maker'. However, it is clear that if wsjr(j) is a correct reading, and Greek Osiris and Aramaic ?(w)sr(y) and Coptic usire/i suggest that it should be (at least, approximately), this sign also had a reading of *ws, or something similar. This *ws looks very much as if it could be related to IE (1.) *wes-, 'stay, dwell, stay over(night)', with several alleged derivatives meaning 'hearth', and 'hearth-goddess' (Latin Vesta and Greek Hestía). Strangely, I can find no reference to this important root in Bomhard's dictionary. This root is probably cognate with Sumerian uš (which I would emend to us₂), 'lie near by, stand, set up (cause to stay[?])'.

We may also notice Dravidian *vis-, 'cook, heat by the sun' (probably better: 'bake').

Now Sumerian engiz, 'cook (n.)', now becomes of interest.

In addition to the Nostratic sequence *w(Vy) becoming g[~]V in Sumerian similar to what we find in Armenian, discussed above, it looks likely as if Nostratic *we also has a similar Sumerian equivalent of g[~]i. The fluctuation among Sumerian s, š, and z is well recognized; and if we emend engiz to **en(-)g[~]is_(x), we have what could very well be the Sumerian response to a Nostratic **wes-, '*heat/bake', a possible basis for Egyptian ws and IE *wes-. Sumerian en, of course, simply means 'master'.

On the basis of the Sumerian data, we may postulate two roots:

PL *FHE-*SHA, 'disperse-stative'='dispersal=glow (with heat), heat'; and

PL *FA-*SHA, 'around-be immobile'='camp around, stay'.

Both of these would have appeared in Egyptian as wz and in IE as *wes-.

This is a depiction of the sign used to indicate st, 'seat, place', and ws in Osiris. I think that most observers would agree that, though this may be a chair of some kind (and it must be admitted that there are depictions of divinities seated on it), it is an unusual chair; and that the prototype may well have been an oven, which it, in my opinion, more strongly resembles than a seat.

If we assume the first root is part of the compound Ws-jrj, which we amend to **Wz-jrj, then the name of Osiris is understandable as 'heat-maker', very appropriate for a god who rejuvenates life, the day, and the year.

This can also be related to IE *wes-, 'spring', the warming time of the year; and *wes(-)peros, 'evening', the time when warmth goes away ([2.] *per-).

The same sign is utilized with t in Egyptian to designate the goddess Isis, which appears in Greek transcription as Isi, in Aramaic as ?sy, and in Coptic as E:se (Sahidic) and E:si (Bohairic). Based on the transmitted forms, the earliest form of the name is almost certainly to be reconstructed as **Jz.t.

If we attempt to follow the logic of the pairing of Isis and Osiris, we are led to the IE root *a/a:s-, 'burn, glow', with derivatives like Hittite haša(-)š, 'stove', which would appear in Egyptian as jz if derived from PL *HHA-*SHA, 'shiny-stative'='shine'.

With this proposal, we can relate this root to Bomhard's:

(381)PN *hhas-/*hh6s- "to burn, to be hot" > PIE *hhes- [*hhas-] "to burn, to be hot"; Afroasiatic: Egyptian HsHs "to burn, to be hot; fire flame" Hss "heat, flame, fire"; PFU *äs3- "to heat, to ignite"

In view of the fact that Hs in Old and Middle Egyptian means 'be cold', it is hardly likely that Budge HsHs and Hss should be given much credence unless what was visualized was the burn accompanying extreme cold, or that some dialect of Egyptian unusually had H as an equivalent to regular j for Nostratic *hh and *?, which is also suggested by Budge's Hs.t, 'seat'. Another possibility is that Hss is a dialectal form of ss, which is Middle Egyptian for 'burn'.

But why would Isis, if it means the 'oven, hearth, stove', or some related concept be written with the sign depicting the unusual seat?

Bomhard's root #434 gives us the answer:

(434)PN *?as^y-/*?6s^y- "to put, to place, to set; to sit, to be seated" > PIE *e/e:s- /*o/o:s- "to put, to place, to set; to sit, to be seated"; PAA *?as^y-/*?6s^y- "to put, to place, to set; to sit, to be seated"; PU *as^ya- "to place, to put, to set"; S aš-te "seat, stool, throne", aš-ti "seat, throne", eš-de, eš-ki "throne"

This root can be analyzed as PL *?A-*SE, 'be-in-contact-with'+singulative='sit down on', which, once again would have resulted in Egyptian jz.

Readers who know Egyptian will be familiar with the interchangeability of s and z from Middle Egyptian times onward. Corresponding to the reading for the versatile Egyptian sign st, we have IE *sed-, 'sit', which we would analyze as PL *SO-*T?O, 'skin-lump='cushion, seat'. The theoretical Sumerian equivalent to Nostratic *sod^w- would be sud; and, in fact, this word exists with the meaning 'sink down', which could perhaps be connected with 'sit down'. The form of the sign with which it is written suggests the possibility of being interpreted as a 'cushion'.

Thus, although I would not dispute a reading of st (Nostratic *sod^w-) for the Egyptian sign, which seems to have so many values, I believe that the readings of js and ws should be emended to **jz and **wz, representing respectively Nostratic *?as^y-, 'sit down on', and *hhasa:-, 'shine', as well as *wasa:-, 'camp around, stay', and *we:sa:-, 'glow, heat'. Accordingly, I make the following revisions and additions:

(434)PN *?as^y- "to sit" > PIE *e/e:s- "to sit"; PAA *?as^y-/*?6s^y- "to put, to place, to set; to sit, to be seated"; Egyptian **jz in **jz.t, "seat"; PU *as^ye- (following Rédei) "to place, to set, to lay, to erect a tent"; S in aš-te "seat, stool, throne"; in aš-ti "seat, throne", eš-de, eš-ki "throne" (source: PL *?A-SE)

(381)PN *hhasa:- "to shine (with heat)" > PIE *a/a:s- "to glow, to burn"; PAA *Has- in Arabic HasHasa "to roast meat on coals"; in Egyptian jz "oven" in Jz.t "Isis"; PFU *äs3- "to heat; very hot; to be very warm"; S as₂ **"roasted", abstracted from as₂-sa₈-še-sa "(sack of) roasted, hulled, emmer" (source: PL *HHA-SHA)

(Addition AS)PN *wasa:- "to camp around, to stay" > PIE (1.) *wes- "to stay, to dwell, to stay overnight"; PAA *Has- in Arabic HasHasa "to roast meat on coals"; in Egyptian wz "**place"; S uš (for us₂) "to lie near by, to stand, to set up" (source: PL *FA-SHA)

(Addition AT)PN *we:sa:- "to glow (with heat), to heat" > PIE (9.) *wes- "to illuminate", discussed under *awes-; *wes- "spring (time of warming)"; in *wes(-)peros "evening (departing warmth)"; PAA *was- in Egyptian **wz "heat" in **Wz(-)jrj "Osiris"; PU probably in *vas^y(-)ke- "ore, copper" (from **vis-[?]); PD *vis- "to cook, to heat by the sun" (source: PL *FHE-SHA)

(Addition AU)PN *sod^w- "seat (cushion); to sit" > PIE *sed-/*sod- "to sit"; PAA *sat- in Egyptian st "seat, place"; S probably **sud "to sink down"; PD possibly in *sut- "to wash (if = "to immerse, to soak")" (source: PL *SO-T?O)

We now return to Bomhard's root #393.

An IE form like *aug-, 'gleam, see', seems to validate our abstraction of IE **aw- as the semantic basis of it and *awes-; however, since we can find this root plus extension only in IE, we have no real way to precisely specify what Nostratic phoneme has resulted in the final IE *s. On the basis of Nostratic *hhasa:- and *we:sa:-, it raises the presumption that the final component is Nostratic -*sa:-, i.e. PL *SHA, stative, yielding 'glowing (with heat)'.

This root cannot be found, at present, for Altaic, Dravidian, or Uralic.

These cognates can be related under PL *HHA-*FHE, 'shiny-disperse'='glow with heat'.

Accordingly, we amend:

(393)PN *hhawe:- "to glow with heat, to shine" > PIE **aw- in *awes- "to illuminate"; in *aug- "to gleam, to see"; PAA **Haw- in Arabic Hawira "to be or become white"; Sumerian u₄ (**ü:₄) "sun, light"; in ug₄ (**ü:g₄) "day, storm (lightning [?])" (source: PL *HHA-FHE)

DISCUSSION: Another important root which is fairly well-distributed is

Here, I believe that Bomhard has captured the essential Nostratic form. I will mention only the Sumerian word munu, 'salt'. If our correspondences are correct, Nostratic *mon- would result in Sumerian **mun. We may first notice that this word is written in Sumerian with a sign that depicts a stake for attachment, inside of which the sign for 'mountain' is written. Presumably we should think of 'that which is attached in the mountains', a rather eloquent way to visualize mined salt. Salt is, of course, eaten; and the common word for 'food' in Sumerian is u₂. In view of its origin, I think it is somewhat probable to analyze munu as mun, '**mountain', combined with u₂, 'food'.

There is also a Uralic word which might be cognate: *mongka-, 'bend', if derived from **mon-ka-.

We therefore reconstruct the root as consisting of PL *MO-*NA, 'flesh-thing'='rounded contour, bulge'.

No suitable Altaic cognate can be presently found.

Accordingly, we amend:

I think it is fairly obvious that this word designated not a sharp range of mountain peaks but rather a contoured hill, gently rising like the bulge of human flesh. Supporting this interpretation of the root is Bomhard's root

I presume by placing this root after #533, Bomhard is inviting the reader to make a connection which I am, for one, willing to make. I would analyze this stem as an extension of Nostratic *mon-, 'bulge', in the more specific sense of 'breast'. Egyptian D represents Nostratic *t?/hsa and *t?/hse; and the element *t?sa, 'body' (PL *T?SA), occasionally occurs with names of bodily parts. Normally, we would expect a response of *dh in IE, however, I have found that IE final *bh, *dh, and *gh, when derived from Nostratic *p?fa, *t?sa, and *k?xa, lose their affrication and consequent aspiration in IE, appearing as *b, *d, and *g. If this is the case here, it would explain the conventionally reconstructed IE form: *mend- (rather than **mendh-).

It will be noticed that Bomhard has reconstructed this root with a palatalized final. Since neither IE nor Egyptian indicate such an origin, he is obviously relying on the Dravidian form *moñci, which, it may be added, cannot be found on Starostin's website in the section for Dravidian roots. Whatever Bomhard's source, I am, however, willing to accept it provisionally with the exception that I believe the form, without the final -*i, was **mont-; and that the -*i has palatalized the preceding two consonants. This is supported by the fact that several of the IE forms are reconstructable as derived from *me/ondi-. Had the Nostratic form been ***monat?s^y-, we should expect IE **mendh-.

Accordingly, we amend:

go to Part Three

go to Nostratic Table of Correspondences ?

PL MORPHOLOGICAL ELEMENTS IN NOSTRATIC

press here to see

to investigate these phonological correspondences in detail, see

TABLE OF PL / IE / AFRASIAN CORRESPONDENCES
TABLE OF PL / IE / ALTAIC CORRESPONDENCES
TABLE OF PL / IE / DRAVIDIAN CORRESPONDENCES
TABLE OF PL / IE / SUMERIAN CORRESPONDENCES
TABLE OF PL / IE / URALIC CORRESPONDENCES

the latest revision of this document can be found at

HTTP://WWW.GEOCITIES.COM/Athens/Forum/2803/NostraticDictionary-2.htm

Patrick C. Ryan * 9115 West 34th Street - Little Rock, AR 72204-4441 * (501)227-9947

PROTO-LANGUAGE@email.msn.com