Peust’s chart overviews the strong compatibility restrictions of pairs of consonants with an asterisk (*) and absolutely no occurrences with (x). His strong restriction means that these pairs of consonants are clearly disfavoured, although they may appear occasionally. I take the occasional appearance of these consonants to mean that they are nonadjacent, again, however, Peust does not indicate any positional variations on consonantal incompatibility. Further, Peust omits the three consonants transcribed as 'a', 'j' and 'n' as he found they were not subject to strong restrictions (1999b:196).
Peust (1999b) does not discuss the general restrictions of these consonants such as articulatory sets or their positioning within a root therefore they will be discussed here supported by Greenberg’s (1950b) and Watson’s (1979) studies into verbal root consonantal compatibility restrictions.
As there are consonants that have been through internal developments, the articulatory divisions are discussed in-depth. I use the standard transcription (put in pointed brackets <> [or '' or "" here] ) along with their posited phonemic representation as put forward in Loprieno (1995:32), which is given for the Old Kingdom period (3000-2000 BCE) and Peust (1999b) for the Late Kingdom period (1300-700 BCE):
The labial series of consonants in Egyptian are:
/b/, 'f' /f/, 'm' /m/, 'p' /p/, ('w' /w/)
The co-occurrence restriction of consonants from the labial group in Egyptian is the most clear and rigorous of all the articulatory series. Although it is seen that the labials can combine freely with the labial glide /w/, Greenberg omitted the labial glide from his discussion because ‘w and y do not consistently pattern with any group of Consonants. It has long been realized that the so-called weak verbs of Semitic, containing w and y in various positions, are ‘rationalizations’ by which older forms containing root u and i were incorporated into the dominant triconsonantal schema’.(1950b:163). Specifically applied to Egyptian this is an instantiation of the “law of Belova” (Takács 1996:355). The initial 'w-' or 'j-' when found in an Egyptian triliteral root are, in many cases, part of the original root of Proto-Afroasiatic with the internal root vocalism *-u- or *-i-, therefore, PAA *C1uC2 > Eg. wC1C2.
These initial glides have previously nearly always been treated as prefixes, and as Watson (1979:100) points out ‘affixal elements do not obey patterning’ (in this root-level co-occurrence restriction, Egyptian is similar to Arabic). Instances can subsequently be seen of the labial glide 'w' /w/ patterning with other consonants from the labial series.
What is interesting from Peust’s chart (fig. 13) is that it shows that the labial glide 'w' /w/ does have strong restrictions against it co-occurring with the velar stop 'k'/k/ and the uvular (?) stop q /q/ .35
However, Watson (1979:105) states that 'w' /w/ does not show any significant patterning in verbal roots and so dismisses any discussion of its co-occurrence restrictions from his paper. Although, when looking at Watson’s root distribution table (1979:101) for first and second root position, in can be seen that the labial glide 'w' /w/ does not pattern with the velar 'k' /k/ or the uvular (?) 'q' /q/ either.36
Furthermore, from Watson’s chart 'w' /w/ is not seen to co-occur with 'p'/p/ and 'f'/f/, although this restriction is not evident from Peust’s chart (perhaps this is due to Peust analysing the nominal and verbal roots, so it can be assumed that /w/ patterns with /p/ and /f/ in nominal forms).
3.2.2 Coronal sonorants
The coronal sonorant series in Egyptian contains the consonants:
Contrary to Peust’s (1999b) findings, that
The articulatory set of coronals is further sub-divided in Egyptian as it is seen there are no co-occurrence restrictions of the stops patterning with the fricatives, although within these sets there are restrictions. This position is reflected in Semitic languages where the same co-occurrence is evident. Importantly there is a ‘rule of transposition’ that is exhibited in Semitic languages whereby the ordering of the consonants coronal stop + coronal fricative > coronal fricative + coronal stop.38
However, Watson (1979 104) states this ‘may have been observed in Egyptian but was not certainly so.’
3.2.4 Coronal fricatives
The class of coronal fricatives is:
's' /s/, 'z' /z/
The co-occurrence of these consonants is strongly disfavoured in Egyptian. Watson (1979:104) lists only two roots containing a co-occurrence of these consonants, but through such a low co-occurrence ‘one may tentatively admit exclusive patterning to have been at work’ (1979:104). Greenberg (1950b:180) also concludes that the co- occurrence of these consonants is ‘very rare’ and only cites one example known to him where they do co-occur. It is noted that in the Middle Egyptian stage of the language these two phonemes merged resulting in only /s/ (Allen 2000:16). It is seen that two distinct graphemes were still used that came to represent the one phoneme /s/ by the Middle and Late Egyptian stages.
3.2.5 Dorsals 39
The dorsal series of consonants are:
In Watson’s study (1979:103), there are absolutely no occurrences of this series of
consonants co-occurring together in the same verbal root, and Greenberg notes ‘I Could discover no instances of Egyptian roots containing two different velars.’(1950b:179). Confusingly though, these two studies label these consonants as velar and post-velar. For Greenberg, the sign he transcribes as 'x' is commonly transcribed as 'x' and he terms this as a ‘post-velar’. Watson gives the Egyptological transcription as 'x', although he follows Greenberg in also terming this sound a post-velar. Loprieno(1995:33) gives the representation of 'x' as a uvular fricative /χ/.
Although Egyptologists are undecided as to whether this sound is thought to represent a velar or uvular fricative, it can be positioned into the dorsal set due to its incompatibility with the other segments in this series. 40
The sign, transcribed as 'q', is thought to be representative of either a uvular stop/velar ejective/labio-velar. Loprieno transcribes 'q' as a uvular stop /q/ (1995:33), whereas Greenberg (1950b) does not define the phonemic transcription of this sign and only posits the Egyptological transcription 'q' (although Greenberg (1950b:180) terms this sound as a velar). Peust (1999b:110) gives the phonemic representation of this sign as a labio-velar /k /. Watson (1979) also only gives a transcription of this sign but for him it is represented as 'q'. 41
Allen (2000:16) states that Egyptian 'q' is ‘A kind of k, probably like Arabic and Hebrew q ...or with some kind of “emphasis,” like q in some Ethiopic languages...’ Moreover, Greenberg (1950b:180) states that ‘the Semitic rules concerning the non-occurrence of velars and post-velars finds its correspondence in Egyptian.’ For clarity, therefore, it is proposed here that the Egyptian velars and ‘post-velars’ should be termed ‘dorsal’. As the term, ‘post-velar’ implies the inclusion into this set of any other sound that is articulated further back than the velar place of articulation (such as the gutturals). This is in line with the Arabic categorisation.
In Egyptian verbal roots, the co-occurrence restriction of the uvular stop/velar ejective/labio-velar 'q' /q/ /k’/ /k /, the velar stops 'g' /g/ and 'k' /k/ or the velar/uvular fricative "x" /x/ /χ/ with each other is upheld.
The gutteral series of consonants are:
Watson (1979:102) describes these consonants as being laryngeals ("a",
He outlines that the combinations of "a" with the other gutturals can be ‘understood as the development of r and l’ (1950b:180). It is evidenced that this phoneme, transcribed as "a", frequently corresponds to Proto-Semitic *r and *l, hence the dual representation given in fig.14f. 44
Watson states that ‘Egyptian A often represents etymological r and l as well as A’ (1979:102). Furthermore he discusses the developments of the other sounds ‘a, likewise, commonly derives from r and less frequently from l besides a itself; and finally H may under certain conditions reflect an original x.’
Conclusively, Watson states that ‘For the time being therefore it must be confessed that no rules of patterning among laryngeals and pharyngeals in Egyptian are immediately apparent and that, in our present state of knowledge no definite conclusions can be drawn.’ (1979:102-103). However, Petráček (1969) finds that 'h' shows incompatibility with 'h' and 'a'. This is also seen from Peust’s chart in fig.13. Interestingly, Rössler (1971) finds that 'a' shows restrictions with the coronal series 'd', 't' and 'z' (also seen in Peust’s chart). 45 As Watson pointed out, this series of Egyptian consonants demand further investigation.46
3.2.7 Ancient Egyptian internal phonemic developments It is evidenced that the series of consonants in fig. 14 have gone through internal developments; this is clearly seen with the analysis of their co-occurrence restrictions with certain articulatory sets. 47
(14) 's', 't', 'd', 'x'
The Egyptian sign transcribed as 's' is proposed by Loprieno (1995:33) to have the phonemic value / /. He states (1995:34) that this phoneme, when palatalised, corresponds etymologically to Afroasiatic *x. 48
This was Greenberg’s theory (1950b:181) although he was unable to support this with any etymologies. Watson (1979:103) shows that 's' ‘does not seem to pattern as though it were a sibilant, and it must be suggested that S behaves as though it were a (prepalatalised) post-velar.’
Although Watson does not discuss its exact phonological nature, he notes that in Old Kingdom writings there is ‘confusion between x and S’ (1979:106). The dorsal nature of this sign is evidenced in Watson’s chart that shows this through the incompatibility of 's' with velars where their co-occurrence is ‘rare.’ From Watson’s analysis (1979:101), this co-occurrence restriction is validated and further 's' patterns frequently with other sibilants, where it has already been discussed that the sibilant series do not pattern with each other. Peust’s chart (fig.13) omits this sign from the compatibility analysis.
The two signs transcribed as 't' and 'd', are given by Loprieno (1995:33) with the phonemic representation of the palatals /t / and /d / respectively (1995:33). 49 However, Greenberg (1950b:180) discusses the fronting of an original 'k' /k/ resulting in 't', and 'd' from a fronted 'g' /g/. Watson terms these sounds as being ‘prepalatalised’ 't' from 'k' and 'd' from 'g' (1979:103). Evidence for their prepalatalisation comes from their incompatibility with the consonants from the dorsal series. Watson’s chart shows that there are no co-occurrences of these two sounds with any consonants from the dorsal series. Greenberg (1950b:180) also sees the incompatibility of these sounds with the dorsal series, ‘It is striking therefore, that there are no verb roots in Egyptian containing both T and a member of the velar stops…it also appears that D does not occur in roots along with a velar stop.’50
A further sign - 'x' is known to be subject to internal developments. Loprieno (1995:33) gives the phonemic transcription of this sign as / / - a palatal fricative. Further, he states that this sound was also, along with 's', the heir of Afroasiatic *x (Afroas. *xanam > Eg. Xnmw “[the ram-god] Khnum” (1995:35)). Watson (1979:103) states that this sound was ‘prepalatalised’. In Watson’s analysis, he finds no instances of roots containing both 'x' and 'x' (velar/uvular fricative). This co-occurrence restriction evidences the prepalatalised nature of this sound. Peust’s chart (fig. 13) also shows that this sound has strong restrictions against its occurrence with the dorsal series of consonants ('x', 'k', although 'q' and 'g' are questioned marked) and interestingly with the three other signs that are subject to internal ‘prepalatalised’ developments ('t', 'd' and 's').
Greenberg’s (1950b:181) study concludes, ‘The general subject of the patterning of consonantal phonemes within the morphemes in Hamito-Semitic languages would seem to be a promising subject of investigation and one whose results must be kept in mind for their bearing on the historical analysis of this family of languages.’ Watson, in supporting the conclusions made in Greenberg’s ‘preliminary attempt’ is unequivocal in his conclusion: ‘...more important however is that the presence of this patterning in Egyptian helps locate Egyptian’s historical position within Hamito-Semitic [Afroasiatic] with slightly more precision that hitherto.’ (1979:105).
.3 Non-Semitic Afroasiatic languages Bender (1978) extended the consonantal compatibility restriction analysis to all the branches of Afroasiatic. Bender found ‘strongly positive results’ for Tamazigt (Berber), an autonomous member of the Northern branch, and the Cushitic languages Beja and Oromo. 51
Further he found ‘More equivocal positive results are obtained for Hausa, Mubi, and Logone (Chadic), Awngi and Sidamo (Cushitic), Welamo (Omotic), Koma (Nilo-Saharan), 52 and Proto-Indoeuropean (all verb roots). Negative results, equivocal or clearcut are obtained for Margi (Chadic), Kefa and Ari (Omotic), Kanuri and Masai (Nilo-Saharan), Proto-Bantu and Moro (Niger-Kordofanian).’ (1978:9). 53
Bender breaks down the consonantal restrictions into their articulatory classes such as labials, dentals (coronals) etc. and gives an overview of their positional incompatibility. Bender concludes that these results obtained show that ‘the co-occurrence restrictions are a good Afroasiatic isomorph, though it is see that Omotic is the weak link, and Chadic is also on the weak side.’ (1978:9-10). 54
The outline given of the consonantal compatibility restrictions evident in these languages primarily shows that these restrictions are not just characteristic of the Semitic language family, but of further language families in the Afroasiatic phylum as a whole. Secondly, that the restrictions are gradient in being (i) positional (adjacency of positions I-II stronger than positions I-III etc), and importantly, (ii) articulatory, as the gradient restrictions always involve the coronal consonantal series.
Source: Kirsty Rowan, Meroitic – an Afroasiatic language?, 2006.
Asar Imhotep puts forth:
The mere argument that because an African term has more than one meaning does not give grounds for dismissal of the obvious connections of Egyptian and the Niger-Congo languages.
Mystery Solver replies:
You simply cannot cherry pick one particular meaning of a word with multiple distinct meanings, out of convenience of the occasion at hand and pass it off as the root context, without understanding the underlying figurative nuance. Specific words are not created out of happenstance, whereby in a single language a specific word is invented many times over and where the meanings attached to each is entirely unconnected to the other.
Asar Imhotep puts forth:
This just goes to prove that linguists has the Egyptian language categorized all wrong in the first place. The reason you have so many definitions for the same word is more than likely it is a tonal language, or, as I have argued, it is an agglutinative language that “hides” various verbs and nouns (just like in the names for the land of Ta-Meri –- Kaa in Km.t and in BaKaa - both meaning Egypt). Ra Un Nefer expressed this point on pg 27 of Metu Neter when he brought out the fact there is no way for to ascertain the meaning of a word in Egyptian verbally unless it was a tonal language. For example the Egyptian word “a-au-au” has the following meanings.
Mystery Solver replies:
Having many contexts for a single term doesn't prove that a language is tonal. For instance, though not perhaps the best examples, in English, black could have many distinct meanings like:
'It was a black day'
'He/she has a black heart'.
'It is a black car'
'He sold the gun at the blackmarket'.
'The blacksheep of the family'
"The room went black"
"Don't let this cloud your judgement."
"He has a black cloud hanging over his head"
"The sky is clear off clouds at the moment"
The word here doesn't need any variation in pitch to be understood within the context it's placed, but more so, by its cotext. It is obvious that these terms have some consistent fugartive undertone, nowithstanding the different contexts.
Asar Imhotep puts forth:
Looking for the "root" of the words will NOT give you ANY indication of the many uses of the word and is why your assumption would not stand up to scrutiny in comparing Egyptian to any other language by the criteria you suggest.
Mystery Solver replies:
If you don't know the root of the word, then how can you just stubbornly go ahead and pick one out of its many distinct applications, and then use that as an argument to suggest that this word derived from so and so language, or that this is the shared common root or genetic link between the words from the host language and a foreign language, while ignoring the other different applications that don't seem to support that position? So of course, you have to have an idea of what the possible root of the word is, before you use it comparative analysis and ultimately, if necessary, reconstruct a proto-term for the proto-language.
Another discussant, Wally, puts forth:
5) Prove that the following expressions are NOT genetically/linguistically related:
EGYPTIAN:Bu nafret su em bu bon, "a state of good has become a state of evil"
WOLOF :Bu rafet mel ni bu bon, "a state of good has become a state of evil"
EGYPTIAN:mer on ef, "he loved"
WOLOF :maar on ef, "he loved passionately"
EGYPTIAN:mer on es, "she loved"
WOLOF :maar on es, "she loved passionately"
EGYPTIAN:mer on sen, "they loved"
WOLOF :maar on sen, "they loved passionately"
Mystery Solver replies:
You have to understand the either “neuter” or “genderless” terms, to be able to recognize their counterparts when they appear in gender suffixed or prefixed pronoun forms, as is the case above, with the term “nfr” - which becomes “nfr.t” [as feminine singular]. While some phonological similarity is obviously invoked in certain terms, like the case is between ‘bw” [Egyptian] and “bu” [Wolof], the “underived” or original application of terms must be examined to see if terms like “bw” [Egyptian], which appear to have multiple meanings, were reinvented multiple times [meaning - phonologically similar terms written in the same letters but without any relationship whatsoever] or simply took upon different disguises, with all ultimately converging on a common ancestral basic theme.
I am aware of ’n’ being used as a preposition, an adjective, a suffixed or dependent pronoun amongst its different applications, but not sure how or whether it relates to the term ‘on‘ used in the examples above, given that we are both familiar with the pronoun Egyptian terms [he, she, they] and the verb [loved] in question. If ‘n’ here does relate to ‘on’, then please clarify its grammatical application in the Wolof counterpart, and also please account for “passionately”, which doesn’t appear in the Mdu Ntr counterpart; for instance, is it denoted by some prefix or suffix [perhaps in the term ‘maar‘] or a lone-standing term [as in ‘on‘]. Whatever the case may be, I have a good hunch that it will weaken the seemingly smooth parallelism in the examples, by not repeating that level of parallelism in the substance behind the terms.
While one cannot rule out some level of relationship between the major African super-language phylums [be it through distant common origins or through historic contacts via immigration, trade networks or conflict], which should at any rate be expected and may well explain some similar terms appearing here and there, such links may or may not be strong when the languages in question are *elaborately* studied.