Pre-Late Egyptian Reconstruction/Egyptian Pronunciation

From Wikiversity
Jump to: navigation, search

Most of this information comes from the masterminds of:

Werner Vycichl
Jozef Vergote
Sir A.H. Gardiner
Pierre Lacau
Antonio Loprieno
Helmut Satzinger
James P. Allen
Carsten Peust

If there's any scholar not mentioned above who partook they will be respectively added in future edits. Thank you for your hard work.


f, p and b

m and n

  • Are pronounced as they are in English: bilabial nasal (m) and alveolar nasal (n).
  • Also appears that the vowels follwing m and n were strongly nasalized culminating in n => velar nasal (ŋ)- sound in sing or possibly uvular nasal (ɴ) at the end of a syllable when in contact by specific consonants (like g, k, ect.), and in other instances (n) and (m) may have had a geminated pro-longed enunciation. There also was a change of the original stressed vowel a into u within the same syllable as (n) and (m) which grew predominance during the Canaanite Vowel Shift.
  • Original < n > develops into < m > in direct contact before the labials < p, m > and sometimes < b > but not < f >[1]. There are also indications that < q > could provoke an assimilation n > m at a Pre-Coptic stage. While < n > frequently assimilates to a following labial, m is not normally affected by a following consonant in its place of articulation, there are a few exceptions in a few words in Coptic but it is rare.
Note: This assimilation is not regular across a morpheme boundary. However the final -ⲚCopt of some proclitic elements may develop into -ⲘCopt before a labial. Assimilation is common (obligatory at least in Sahidic)[2].

The Four H's: h, ḥ, ḫ, h[3]

Egyptian < ḫ > frequently substitutes Semitic /x/ in New Kingdom loans.
Occasionally used sporadically/irregularly in lieu of a voiced pharyngeal fricative (ʕ) from Semitic loan words[4].
This sound has two (Pre-)Coptic descendants: and h=> which is palatalized. This occurred in the New Kingdom with ḫj (h̭) occasionally used for older ḫ where the Coptic descendant is h . This new digraph in the New Kingdom indicates a secondary palatalization (x > x).
  • h = Has an interesting history. In the earliest time there was no phoneme corresponding to /h/. The sound /x/ corresponding to < š > was frequently palatalized to /sh/ in the Old Kingdom. One of the most accepted hypotheses of this sound is that the grapheme < š > which formerly had expressed a back fricative thus became primarily a means of writing a palatal fricative. In the minority of words in which /x/ escaped palatalization, the preserved back fricative now had to be expressed by a different sign for which /h/ was invented[5].
Generally believed to have been a voiceless palatal fricative (ç) ... It appears most scholars believe this sound (agreed upon that it was an < h > of some sort) had a hint of palatalization associated with its articulation, especially during the Middle Kingdom and onward. Other scholars claim this sound was initially a voiceless counterpart of < ḫ > which then, < ḫ > would have needed to have been voiced- in this final scheme of things, ḫ (excluding the hypothesis of the split ḫ ~ h) was palatalized to < š > under certain environments.
The conditions for the palatalization of š > /x/ are not known. There is perhaps a connection to the palatalization of "back stops" discussed in the section below called Palatalization under Palatalization caused by labials in the Old Kingdom[6].


  • voiceless palato-alveolar sibilant (ʃ). The consonants transcribed as h and š are regularly distinguished from one another only after the Old Kingdom. Earlier words that later have h are regularly spelled š (but not vice versa).



k, q and g

d and t[7]

  • d may have been an ejective apical consonant (t̺ˈ) or apical non-ejective d̺. Other hypotheses are that it is either laminal denti-alveolar (d̪) similar to Spanish, or like emphatic/pharyngeal of Arabic. It's actual pronunciation is a matter of debate but it may have been something similar to an ejective apical t̺ˈ or apical non-ejective d̺ especially due to ejective/aspiration alternations with consonants and the change from orthographic d into t in Coptic.
  • t = (if d ~ t̺ˈ ) may have been the non-ejective aspiratable counterpart of d, thus: t̺(ʰ).

d and t

r and l

  • r = is a central alveolar flap (ɾ), almost identical to Italian or Spanish r or the dd in English ladder.
  • Although < l > did not have an orthographic letter in Egyptian until the New Kingdom/Late Egyptian, the lateral approximate (l) did exist and may have sounded something like the Japanese alveolar lateral flap (ɺ)[8] or less likely retroflex flap (ɽ)[9].
  • The sound < l >, at least when it was a loanword or borrowed from another foreign language into Egyptian at such point it was now an Egyptian word, was obviously confused and could have quite possibly been pronounced between an < r > and an < l > or any other random consonant without rhyme or reason. The sound did swing one way or another between dialects which played a major role in the ambiguity of its approximate pronunciation. The hieroglyphics are not of much help as this is the primary reason why a thorough explanation can not be had- which proves, at the least, that there was confusion in articulation of the sound once rendered in writing. The sound of < r > was much more stable if it was borrowed from a foreign source[10].

Glides & Weak Consonants[edit]

  • The Egyptian orthography < ꜥ > may have originally been a voiced dental apical non-ejective /d̺/[11], ejective /t̪ʼ/, or emphatic/pharyngeal /tˤ, dˤ/ in the Old Kingdom - (at the least they may have existed in a dialect or two). Then eventually evolving into a glottal stop < ʔ > (if the original sound was ejective or a voiced dental) or a pharyngeal < ˤ , ʕ > (if the original consonant was pharyngeal)- this was achieved by the time of Middle Kingdom[12].[13]
  • After < ꜥ > adopts the pharyngeal sound value < ˁ > or glottal stop < ʔ >, there is a tendency for it to undergo dissimilation to < j > in the neighborhood of < ḥ >[14]:
ꜥḥ ~ jḥMEg - palace.
mḥꜥw ~ mḥjwMEg - flax
  • When < ꜥ > adopted the value < ˁ > or < ʔ > at about the time of the Middle Kingdom, it could be dissimulated to < j > in the neighborhood of < ḫ > (perhaps h), as it was in the neighborhood of < ḥ >[15].
  • This phoneme was generally believed to be The Egyptian Ayin a voiced pharyngeal fricative (ʕ), possibly a voiced epiglottal (pharyngeal) central approximate (ʕ̞) but this has slowly been disregarded.
  • Egyptian < ꜥ > was occasionally devoiced to /ħ/ and then retained in Coptic as Ϩ. This development is found primarily in Bohairic, but is by no means regular[16]

Note - There is a tendency for < n > to replace < Ꜣ > in the hieroglyphics in some roots. It is unclear if at times < n > represented /n/ or /l/ because < n > could also be used in lieu of /l/; and /r/ and /l/ were also interchanged orthographically as well as phonetically. In such case the verb MꜢꜢ - to see was written MꜢn.f - he sees in some pronominal forms. MꜢꜢ appears to come from the pro-Afro-asiatic root mVrVʔ[18] which indirectly shows that some sort of metathesis and assimilation occurred with irregular omission of < r > in the proximity of /ʔ/[19]. With this being said, it is my hypothesis that this sequence would have been initially pronounced mǎʁ-ləf possibly merging into ʔ-ləf => mꜢn.f - he sees.[20]. There is also another verb with a similar hieroglyphic spelling which follows an identical scheme in Coptic: ⲘⲈ[21] - love = ⲘⲈⲢⲈ- , ⲘⲈⲢⲒⲦ=.
There are also many times when /r/, /l/ and < j > replace < Ꜣ > indirectly showing us that < Ꜣ > was merely a graphical substitution for those weak consonants when not fully enunciated in a word during the intermediate stages of the Egyptian language.
  • By the Middle Kingdom, < ɜ > turned into a glottal stop (ʔ)- which is similar to the gulping/choking sound found in the British pronunciation of bottle.
  • Was originally regarded as The Egyptian Aleph < ا , ʔ The Glottal Stop> but the glottal stop enunciation of this phoneme appeared to exist as a secondary feature during the Middle Kingdom and because of this experts have come up with new hypotheses.


  • Either represents y in yes or can be used in lieu of j. It appears to have been originally used for Afroasiatic y in loan words or group writing but also took on the unstable characteristics of Egyptian < j >.


  • Since prehistoric times |j| is believed to have represented a palatal approximate (j, ʲ).
  • By the beginning of the Middle Kingdom |j| assimilated into a glottal stop (ʔ, ʔ) before an unstressed vowel in initial position (jawīn > ʔawīn), as well as at the end of a syllable of a stressed syllable (bājnat > bāʔnat).
  • In Sahidic Coptic, the vowel version is always spelled < ⲈⲒ > in initial position; ⲈⲒⲚⲈ ~ (íne), ⲈⲒⲤ ~ (ís). Internally or in final position both < Ⲓ > and < ⲈⲒ > are used but < Ⲓ > is preferred.
  • In Sahidic Coptic, the combination < ⲈⲒ > less commonly < ⲈⲈⲒ > can indicate two separate letters, a knowledge of the grammar is needed to distinguish between the two uses, for example, ⲈⲒ has the value Ⲉ (vowel) + Ⲓ (semi-consonant):
1) in the demonstrative adjectives: ⲠⲈⲒ, TⲈⲒ, ⲚⲈⲒ ... ⲠⲈⲒⲢⲰⲘⲈ (pey-rōme)... same is true of ⲠⲀⲒ, TⲀⲒ, ⲚⲀⲒ
2) in the first person verbal prefixes of the forms -ⲈⲒ, -ⲚⲈⲒ, -ⲘⲈⲒ
3) and in a few isolated words like; ⲈⲒⲈ (eye) - a participle, ⲈⲒ (ey) - come
  • There are specific rules in how (and when) this sound was pronounced according to Coptic[22] [the same can be assumed in the earlier stages of the language with some modifications[23]):
  • Glides which directly precede the stressed vowel are normally preserved in Coptic (and pronounced in Egyptian):
ⲈⲒOⲘ (ym) - sea
ⲒⲈⲒⲢⲈ (jr.t) - eye[24]
But <j>/<ɜ> can also be lost. This concerns the following cases:
1) In most verbs <j> / <ɜ> is lost even before the stressed vowel[25]
2) <j> is usually lost before CoptⲎ (rule by Vycichl 1940: 87)
jꜢw.t - cattle, conserved in ⲦⲂⲚ-ⲎS., ⲦⲈⲂⲚ-ⲎB. - animal
jp.t - number ~ ⲎⲠⲈS., ⲎⲠⲒB.
jrp - wine ~ ⲎⲢⲠS.B.
An exception is Ꜣbw - Elephantine (name of a town) > ⲒⲎⲂS.. Note also rmy.t - tear (since Middle Kingdom) > ⲢⲘⲈⲒⲎ but ⲈⲢⲘⲎB.
3) There's a few others in Coptic with unexplained reasons for the loss of the glide.
  • A glide is usually retained if it directly follows the stressed vowel[26]:
OⲨϪⲀⲒ [widǐɜ] - to be healthy
ϨⲀⲒ [hy] - husband
  • Roots with medial <j> / <ɜ> eventually conform to a hollow root in Coptic. In the older language there may have been other rules applied determining when to simplify the pronunciation/spelling, until now, there is not enough research but it can be thus assumed, when a root was used as a verb, medial <j> / <ɜ> was most likely omitted (or modified) because of constant fluctuations due to the additions of suffixes and nouns (or any verbal form being used as a noun) used the full pronunciation. For more examples and explanations, I have dedicated a page to the hollow verb here.[27]
  • <j> / <ɜ> is always lost in pretonic position[28]:
ⲀⲚOⲔ - independent 1st singular pronoun ... ect.
  • In hieroglyphs when used as a clitic/suffix/addition at the end of a word (or beginning of a word) it may have signaled a vowel rather than Vj or jV.


  • Is like it is in English labio-velar approximant (w).
  • Like Egyptian -j, w may have signaled a vowel rather than -Vw / -wV especially when used as a suffix/clitic/addition.
  • The diphthong -aw may have sounded like äʊ or äʊ̯ which is a popular pronunciation used in many Afroasiatic languages.
  • <w> is never lost before the stressed vowel.
  • <w> can be preserved or lost in pretonic position. The conditions are not clear. Sometimes there is dialectal variation in Coptic[29].
  • If a cluster of two glides directly follows the stressed vowel, only one of them is preserved in Coptic. Two different developments are possible for a cluster of two glides one of which is <w>. There appears to be no rule to predict which development applied[30] and the Coptic dialects may show different developments:
1) The first possibility is that Coptic has /w/, no matter whether <w> appears in the first or in the second position of the cluster in Egyptian. A good explanation would be that assimilation took place at some stage of the language:
jry.w - fellows ~ ⲈⲢⲎⲨS., ⲈⲢⲎOⲨB. ~ jarjǔw?!
gꜢw - narrow ~ ϬⲰOⲨ ~ gāꜢaw
gwꜢ - to push off ~ ϬⲰOⲨS., ϪⲰOⲨⲚB. ~ gāwaꜢ
tꜢw - wind ~ ⲦⲎⲨS., ⲐⲎOⲨB. ~ tǔꜢaw
tꜢ.w - to be taken (stative) ~ ϪⲎⲨS., ϬⲎOⲨB. ~ tǔꜢaw
djw - five ~ ϮOⲨS., ⲦⲒOⲨB. ~ dǐjaw
Note: The combination Ꜣw/jw usually renders OⲨCopt
2) Another possibility is that the first consonant survives in Coptic. Thus, if <w> is the first consonant, it is preserved:
nǐwiꜢ - to see ~ ⲚⲀⲨ
ḥāwaꜢ - to be rotten ~ ϨOOⲨS., ϨⲰOⲨB.
3) If <Ꜣ> is the first consonant, it can be preserved or lost in Coptic, as is true for word-final <Ꜣ>:
ḥkꜢw - magician ~ ϨⲀⲔOS., ⲀⲔⲰB.
ḥfꜢy.w - snakes ~ ϨⲂOⲨⲒS.B.
sbꜢ.w - discipline ~ ⲤⲂOⲨⲒS.B.
zꜢw - beam of wood (since Middle Kingdom) ~ ⲤOⲒS.B.
zꜢw - a town (Sais) ~ ⲤⲀⲒS., ⲤⲀB. - attested as sa-a-a (probably to be interpreted [sajɘ] or similarly) in Neo-Assyrian cuneiform transcription (see Borger 1996: 20)
Note:[31] If the preceding consonant was lost, the glide comes in direct contact with the stressed vowel. Therefore it preserves its consonantal nature in Coptic, however <j> is lost in the Sahidic dialect (on the phonological interpretation of the Bohairic forms):
jrj.j < jrj.w - (he) is done (3rd pers masc stative) ~ OS., OⲒB., ⲀⲒF.
mrj.t - to love ~ ⲘⲈS., ⲘⲈⲒB., ⲘⲈⲒⲈA.
hꜢj.t - to fall ~ ϨⲈS., ϨⲈⲒB., ϨⲈⲈⲒA.
r-prj.t - temple ~ ⲢⲠⲈS>, ⲈⲢΦⲈⲒB., ⲢⲠⲈⲈⲒA.; ⲀⲢⲠⲀⲈ / ⲀⲢΨⲀⲈOld Nubian - borrowed from the Bohairic form.
mꜢwt - to think (since Middle Kingdom) ~ ⲘⲈⲈⲨⲈS.A., ⲘⲈⲨⲈB.
tbw.t - sandals ~ ⲦOOⲨⲈS.A., ⲐⲰOⲨⲒB. - with irregular loss of [ b ].
Note II:If the preceding consonant was not lost, then it directly precedes the glide in Coptic. There is thus a tendency for the glide to be vocalized. <w> is preserved as a consonant in some dialects of Coptic, and vocalized to /u/ in others. <j> is always vocalized; however in most varieties of Coptic it does not develop into /i/ - as might be expected -, but it merges with /Ⲉ/. Only in certain varieties of Lycopolitan is it indeed written < Ⲓ > and thus kept distinct from the final vowel /ə/ that is written < Ⲉ >.
rsw.t - dream ~ ⲢⲀⲤOⲨS. [rasu], ⲢⲈⲤOⲨA., ⲢⲀⲤOⲨⲒB. [raswi], ⲖⲈⲤⲂⲒF.
mtw.t - poison ~ ⲘⲀⲦOⲨS. [matu], ⲘⲈⲦOⲨA., ⲘⲀⲐOⲨⲒB. [matwi], ⲘⲈⲦⲂⲒF.
pḥ.wj - back ~ ⲠⲀϨOⲨS.A. [pahu], ΦⲀϨOⲨB.
šnw.t- granary ~ ϢⲈⲨⲚⲒB. [shewni] with metathesis.
tzj.j > tzj.w ( < w > is only graphical) - to be exalted (stative) ~ ϪⲀⲤⲒL., ϪOⲤⲒS.
mnjw - herdsman ~ ⲘⲀⲚⲈS., should be ⲘⲀⲚⲒ in in Lycopolitan which is accidentally not attested.
3) Some words are attested with both developments which may be distributed over different dialects:
šmꜢ.w - aliens ~ ϢⲘⲘOⲈⲒS., ϢⲘⲘⲰOⲨB.; But -ⲰOⲨB. may also be a secondary development[32].
gꜢw - to be narrow; the infinitive develops into ϬⲰOⲨS. and the participle form is ⲔOⲨⲒS. - narrow.
The pluralic adjective ꜥꜢ.w - great, is preserved as the final component of certain compounds as follows:
ϨⲖⲖ-OⲈⲒS., ⳊⲈⲖⲖ-OⲒB., ⳈⲖⲖ-ⲀⲒA. - old men
4) We have an irregular development in
rwj - to depart, go away ~ ⲖOS.A2, ⲖⲀS.F. ... Carsten Peust states: It is unclear whether stress was originally on the first syllable (ra'wVj) or on the second (rV'waj). The vowel class can be explained in both cases, but the loss of <w> is irregular in any case... There are several other words which follow a similar pattern:
zwr (> zjr) - to drink ~ ⲤⲰS.B., ⲤOⲨS.A.A2
nꜥy.j (> nꜥy.w) - to go (stative) ~ ⲚⲀ
5) If the cluster contains only <Ꜣ> or <j>, then Coptic has /j/:
mꜢj - lion ~ ⲘOⲨⲒS.B. ( = muj)
ḥꜢ.ty-j - my heart ~ ϨⲦⲎⲒ
The following example is similar, but /j/ is lost in Sahidic:
dy.j - (he) is given (3rd pers. sg. masc. stative) > dy.w (w> is only graphical), ⲦOS., ⲦOⲒB. , ⲦOⲒⲈL., ⲦⲈⲒⲈA..

Elimination of Weak Consonants[edit]

  • /t, r, j, w, ꜥ and Ꜣ/ are deemed The Weak Consonants and usually will fall away at the end of an unstressed syllable[33]:
natǎrat ~ natǎraʔ
  • /t, r, ꜥ and Ꜣ/ can additionally also fall away even in stressed position:
jātraw ~ jāʔraw
  • /w/ is the most stable of the weak consonants appearing in most positions.
  • Sometimes other consonants such as /d, t/ (especially when assimilated to /t/ in later phases of the language), /m/, /n/, /l/ & periodically /b, p, k, ect.../ can all fall away in an inconsistent manner.
  • These phonological processes are not always indicated in the hieroglyphics but is majorly used in Coptic and to a lesser degree in Demotic.

Emphatic/Pharyngeal VS Ejective/Glottal[edit]

There is some debate as to whether q, g, d, z, t, and d were emphatic, pharyngeal, glottal or ejective.

Emphatic groupings (a fancy way of implying pharyngeal characteristics /tˤ/) tend to be the most accepted theory in the field with glottalized groupings (a fancy way of implying ejective features /tˀ/) not far behind. In Egyptological sources, both emphatic and glottalized consonants are not entirely defined and instead can imply a much more broad spectrum of phonological features.

I have chosen to lean towards an ejective grouping of consonants, mainly because other Egyptian consonants were ejective in nature which causes speculation as to other consonants were as well and that these series of consonants constituted an ejective grouping- there is some indirect evidence in Bohairic orthography. Ejective consonants are also fairly prominent among other Northern African languages near to Egypt and extend down the Nile river into the Horn of Africa.

This diacritic /tʼ/, in IPA, can either represent Emphatic or Glottalized.


Under specific environments velar consonants shifted to their palatalized counterparts.

Original Modified
k t [34]
g, q d [35]
ḫ and h š

In Pre-historic Egyptian, those roots which appear to be an Afroasiatic (better analyzed as a Semitic) cognate which contained the vowel /i/ following the Afroasiatic consonant(s) /k, g/ sometimes /q/ (and some other consonants) shifted to their palatalized counterparts in Egyptian:

-ki (2nd fem sing pronoun - you) = -at(i)

Former /s/ (including /z/) frequently develops into a palatal sibilant < Ϣ > in Coptic by assimilation to another palatal in the word. The principal rules are the following[36]:

  • /s/ usually becomes ϢCopt if there is another Ϣ (< š) somewhere in the same word[37]:
zšn (possibly original zššn) - lotus ~ ϢⲰϢⲈⲚ
  • A palatal sibilant Ϣ which is derived from ḫ causes assimilation in Bohairic only[38], OⲨⲰⲤⳈA.:
ḫsf - to repulse ~ ⲤⲰϢϤS. (with metathesis), ⳈⲰⲤϤA., ϢⲰϢϤB.
  • If an original palatal stop (i.e. Ϭ / Ϫ < d, t) follows somewhere in the same word, /s/ is assimilated, though not always in all dialects:
sdd - to talk ~ ϢⲀϪⲈS., ϢⲈϪⲈA., ⲤⲀϪⲈB.
If it precedes, there is no assimilation:
tzj - ϪⲒⲤⲈS.A., ϬⲒⲤⲒB.
  • A palatal stop derived from g / k has an even less palatalizing effect. It can only affect < s > in direct contact:
sg - stupid ~ ⲤOϬS.
  • The palatalization of /s/ can occasionally already be observed in Demotic. In words of this type, palatal assimilation can be observed even earlier[39].

It is generally believed that Egyptian did not originally have palatalized consonants, but that they appeared under specific environments and continued to be used this way into Coptic sometimes with slight instabilities.

Palatalization caused by labials in the Old Kingdom[edit]

  • /k/ and /q/ are incompatible with /p/ and /w/ [40] (and perhaps other labials, for example /b/ ) within an Egyptian root and also shifted to their palatal counterparts.
kbw - sandals ~ tbw
kw (normally tw) - dependent pronoun 2nd pers. sg. masc.[41]
tp - to load, carry
wꜢd - green, yellow


Vowels for Egyptian are fairly difficult to reconstruct since we don't have any credible known examples. In studies throughout the field there appears to be two major theories in regard to the Egyptian vowels:

  • 1) the vowel was always |a| and due to the proximity of specific consonants /a/ morphed into another vowel causing harmonious enunciation[42], or
  • 2) the Semitic-centric vowel inventory of |a-i-u| existed since the Prehistoric Egyptian times.

The following may be applicable using either of the two theories:



  • near-close near-front unrounded vowel - like English sit. In some phonetic environments, for example at the end of a word, may have been pronounced like 'ee' in 'meet' which is also prominent in Egyptian Arabic.
  • In early New Kingdom, short stressed /i/ turns into short stressed /e/.
  • At a later date, |i > e| and short stressed /u/ merged into /e/[43].

ʊ (ɯ̽, or ɯ)

  • Originally /u/ may have been pronounced as a close back rounded vowel (u) in all positions.
  • near-close near-back vowel (ʊ) may have been used for short /u/ especially in diphthongs, as this is a very typical pronunciation in most North African languages. /ʊ/ is also an allophone of short /u/ in colloquial Egyptian Arabic.
  • At some point between the Middle Kingdom and Late Kingdom long stressed u may have been pronounced as a close back unrounded vowel (ɯ) due to the speakers' adaptability to enunciate (u) as an unrounded vowel in relation to the vowels /a, i/ - that is articulating /u/ with the lips unrounded rather than pressed together, just like one would pronounce /a/ or /i/[44].
  • During the Late New Kingdom long stressed /u/ turns into long stressed /e/.
  • It is evident that /u/ went under some sort of crisis which can not yet fully be explained because CoptOⲨ becomes a rare vowel (when it is not used as a consonant or in lieu of long stressed /a/ following a nasal). The original Pre-Coptic Egyptian vowels /u and i/ also appear to be indistinguishable from CoptⲎ. And the vowel /u/, whether from an original /u/ or from CoptⲎ, is used sparingly in infinitive vocalizations.

Unstressed Syllables[edit]

  • I would assume, based upon other Afroasiatic languages, that in Ancient Egyptian speech unstressed /a/ was probably reduced to /ə (schwa)/ or stayed the same, unstressed /i/ probably reduced to open-mid front unrounded vowel (ɛ) or a close central unrounded vowel (ɨ) or stayed the same and unstressed /u/ probably reduced to /ɛ, i, ɨ, ə/ or stayed the same. Eventually unstressed syllables became even more reduced [or deleted within syllables] through into Coptic, a similar phenomena exists with the Berber languages. In Coptic orthography, unstressed vowels (and sometimes even stressed vowels) have no straight forward pattern among them (though maybe more research in this area may prove otherwise), but thus far we can analyze and cross-examine within the Egyptian language as a whole and compare with all other sister languages.
  • Original -Vw, -Vj/y, -Vr, -Vt, ect or any other possible combination sometimes have been entirely reduced to /ɛ, i, ə (schwa),ɨ/ in unstressed syllable position.


Egyptian appears to have had a distinct stress pattern culminating in various vocalizations if stress is relocated (according to Coptic spellings- one must thus adduce that this also occurred in the earlier phases of the language[45]). Every stem in Egyptian has stress on the penultimate syllable (that is second to last syllable) and under two main circumstances the stress moves and may cause a vowel change (or change in vowel length) on the stressed syllable and abridged vocal variations (or vowel deletion) in unstressed syllables. These following rules apply to the Egyptian language as a whole but is best learned specifically with the Coptic infinitive + subject/object additions:

  • Additions of various prefixes, suffixes (and words which act like prefixes/suffixes) or some kind of auxiliary (this would be best compared to English words like: reiterate, bio-organism, ect). For example:

In the case of a nominal object, the addition of the object directly to the verb causes the tone to pass from the verb to the object, A.Ϥ.ϩETB.Π.PωME. We might call this object the ‘tonal object’. However, when the object is a pronoun, it does not itself receive the tone, but rather follows the stressed syllable in the pronominal form of the verb; e.g. A.Ϥ.KOTB.Ϥ. It might be described as the ‘post-tonal object’, but in view of the fact that some verbs, owing to the loss of original consonants, do show a tonal stress on some suffix endings (e.g. MECTω.K - To hate thee, CAϩω.Ϥ - To set him up, etc.), a better name would be ‘direct suffix object’.
E.g. A.N.PωME MEPE.Π.KAKE N.ϩOYO E.N.OYOEIN - Men loved darkness more than light.

A.Ϥ.NOϪ.Ϥ.E.ΠE.ϢTEKO - He cast him into prison [46].
  • When two or more words are placed closely together to form a compound noun or group, the tone falls on the last word only and the formative vowel of the preceding word or words shortens; e.g. ϩOY-MICE ‘birthday’ (from ϩOOY ‘day’ and MICE ‘to give birth to’), PEI.PωME ‘this man’ (from ΠAI ‘this’ and PωME ‘man’), CKPKP.Π.KOT 'to revolve the wheel’ (from CKOPKP ‘To roll’ and Π.KOT ‘the wheel').

Plumley [47] mentions, 'the older forms of the language show that, apart from some verbs mentioned above, originally the direct object— either tonal object or direct suffix object— was the normal usage with all tenses'.

James P. Allen [48] mentions, The rule of Egyptian word order is: VsdoSOA (Verb, subject pronoun, dative (ex: n = of), object pronoun, Subject, Object, Adverbs and prepositional phrases) = capital letters refer to nouns, and small letters refer to pronouns. Pronominal datives (d) and objects (o) are separate words but were probably pronounced together with the verb without a separate stress of their own. Thus:

rdjt n.f t-ḥd (giving him white-bread) probably had two stresses, rdjt-n.f and t-ḥd, just like in English: GIVing-him WHITE-bread.
rdjt.k n.j (n)syt.k (your giving me your kingship) probably had one stress on rdjt.k-n.j and (n)syt.k, just like in English: your-GIVing-me your-KINGship.
jrt n.f st (to do it for him) probably had only one stress, just like in English: to DO it for him.

Nominal subjects, objects, and datives, on the other hand, tend to be stressed separately:
rdjt mntw tɜwj n jtj (MONtu’s GIVing the-Two-LANDS to-the-SIRE).

It is also to be noted, in relation to subject/object additions, that in Egyptian the subject and object are behind the verb and in Coptic the subject is in front of the verb:

prt sm - the emerging of the sem-priest
prt.s - it's emerging

Π.PωME CωTM - the man hears
Ϥ.CωTM - he hears

And the direct object follows the subject in Egyptian, and in Coptic the direct object follows the verb:

jrt jst jɜkb - Isis's making mourning
tzt.j jb.j - my lifting up my heart
rdjt.f wj m ḥɜt hrdw.f - his placing me in front of his children

†.NA.Bωλ - I will loose => †.NA.Bλ.THYTN - I will loose you [uses the pronomial form]
A.Ϥ.ϩETB.Π.PωME - He killed the man [with a nominal object the construct form is used]

This results in different stress patterns between Egyptian and Coptic spelling.

Antonio Loprieno[49] mentions: 'Infinitives may be used in construct or in pronominal state followed by the subject (with intransitive verbs: pr.t=k /'pirtvk/ "your going forth") or by the object (with transitive verbs: sdm=f /'saɟmvf/ "to hear him"; the subject is introduced in this case by the preposition jn).

Loprieno also gives an example of a noun augmented to a nominal verb stem (which he terms emphatic/nominal [like its Semitic equivalent iparras] - this pattern follows the a-Type vocalization)[50]. This inflectional verb pattern is basically identical to that of the Egyptian infinitive:
sdm zɜ=j ~ saɟam'zi:ʀaj - my son listens
jrr=s ~ ja'ra:rvs - she does

Pre-Late Egyptian Construct Form[edit]

In order to render a shift in stress-tone orthographically for Pre-Late Egyptian, I have chosen to use an acute accent which goes on top of the secondary stressed vowel. For example:

prt sm => pǎratOEg / pǐritMEg + sǎm (pírit-sǎm) - the emerging of the sm-priest


The instability of vowel qualities within the Egyptian language may have had two main explanations:

  • Some type of language contact or multilingual borrowings with a Nilo-Saharan language (or more than one as well as any other unrelated Afroasiatic language/s) in Egyptian's pre-history. Nilo-Saharan languages utilize distinct linguistic features such as advanced retracted tongue root (ATR/RTR) as well as vowel harmony and tone/pitch possibly explaining some of the vowel inconsistencies in Egyptian. These linguistic features were lost in Egyptian Pre-history but remnants may have still existed (for example relocation of tone-stress occasionally changes vowel qualities in Coptic- this is not too common in most Afroasiatic languages (other than Hebrew) but is reminiscent of Nilo-Saharan pitch-tone relocation and/or vowel harmony). This also seems to be the case with Cushitic languages which also have Nilo-Saharan/Niger-Congo sub-phylum characteristics.
  • The infamous Canaanite shift contributed to a total reorganization of vowel qualities before the turn of the century.

.... ..... ..... ......

  1. Carsten Pesut pg 161.
  2. Carsten Peust pg 161.
  3. This article gives more insight on some of these sounds:
  5. Carsten Peust pg 116.
  6. Carsten Peust pg 116.
  7. t colloquially was not pronounced when ending a syllable, for ex: nāf(i)rat => nāfra, nāfrə or nāfre.... Note: stressed -a ~ u did not occur until the Late Kingdom when 'stressed -a ~ o.
  8. Merely a hypothesis based upon Coptic Fayyumic, in which case would be better regarded as a dialectal variation of |r| possibly of < l > in the main Sahidic or Bohairic dialects.
  9. Merely a hypothesis based upon Coptic Fayyumic, in which case would be better regarded as a dialectal variation of Coptic orthographic |r and l|.
  10. This < r ~ l > phenominon exists in several languages, for example in the Spanish dialect of Puerto Rico:
    Note: Taking into account that it is believed Spanish < r > and Egyptian < r > are articulated identical only showing that a lateral tap < r > can sound like an < l > with no known genetic relationship between the two languages.
    Lateralization of /r/ to [l]
    The /r/ that comes at the end of a syllable (not followed by a vowel) is often changed to an /l/, so that words like "perdón" (forgiveness) and "Puerto Rico" become "peldón" and "Puelto Rico" respectively.
  11. This is currently the most widespread hypothesis. This is also a sound that is lacking from the Egyptian orthographic inventory and alphabet. The non-existance of this sound in Egyptian tends to be considered highly unusual to experts.
  12. The general belief of this theory is if /d/, /d/ and /t/ were pharyngeal(ized) the original consonant was /d/, /d/ or /t/ possibly emphatic /t/. It appears a dialect (or a variety of different dialects) played a crucial role in the dubious usage of this letter in the Old Kingdom as can be seen in comparison within the hieroglyphs. The same can be said if /t/ was ejective, eventually being assimilated into a glottal stop by the Middle Kingdom.
  13. In the lexicon of the OK, Egn. < ꜥ > is incompatible with dentals/alveolars, in particular with < d > and < z >. There are no roots with ꜥd(...), ꜥCd, dꜥ(...), dCꜥ, Cꜥd, Cdꜥ; ꜥz(...), ꜥCz, zꜥ(...), zCꜥ, *Cꜥz, *Czꜥ. This proves that < ꜥ > it was – originally – a dental/alveolar itself....
    It is compatible with laryngeals ( ḫ, ḥ, h, š [< h]) and with all velars/palatals except k / t, hence g, q, d [< ḳ]. This proves that Egn. < ˁ > was – originally – not a laryngeal itself... pg 7.
  14. Carsten Peust pg 103.
  15. Carsten Peust pg 104.
  16. Peust pg 105.
  17. In my opinion, < Ꜣ > can thus primarily be regarded as some kind of guttural r sound which fell away in pronunciation at an early time period. It appeared that as the language progressed there was a gradual reduction of guttural sounds.
  19. One rule is apparent, though, namely that <r> is always reflected as Coptic Ⲗ in a word which also contains Ꜣ ... Peust pg 129.
  20. There are a few instances of irregularities in consonants emerging in pronominal forms, for example: (excluding -t of weak verbs) ... ⲤϩⲀⲒ - write = ⲤϩⲀⲒⲤ- or ⲤϩⲀⲒⲦ-.
  21. In my own research I have come across the stem for 'love' in Akkadian: rāmī a cognate of Hebrew: raHǎmíy shown in the Hebrew biblical conjugated form: ʔɛ|rHɔm|əkɔ́ʔ - I love you, taken from this source:'to%20love'&f=false
    This shows a possible root of rV(HV)Vm(V) - love [or the like] with some sort of metathesis in Egyptian following a similar pattern to mVrVʔ - see.
  22. The following rules are taken from Carsten Peust's book pgs 143 and on....
  23. We can observe that morphophonological alternation is found only in late words which are first attested in the New Kingdom. The explanation seems to be that morphophonological alternation in more ancient words was eliminated early by analogical levelling... Carsten Peust's book pg 143.
  24. Proto-Afro-Asiatic: ʔir - eye...
  25. The importance of the morphological category suggests that analogical levelling has taken place here. Fecht (1960: XIVf.) provides a plausible explanation that many verbal forms of the suffix conjugation, which was very familiar to the older language before it became almost extinct in Coptic, had suffixal stress. /j/ would thus have been lost regularly here in pretonic position, whence the /j/-less variety was extended by analogy to forms bearing initial stress such as the infinitive.
  26. However, there are several cases of loss which remain unexplained... -OCopt [ꜥǎꜢMEg] - to be great ... ect. Peust pg 145.
  27. Carsten Peist pg 148, states: ... It was discussed in § 3.13.3 that in the succession stressed vowel - <w> - consonant, <w> was usually written <y> from the 1st Intermediate Period on. This explains why <w> in this position is not usually preserved in Coptic. But <j>, including <j> <Ꜣ>, is not preserved in Coptic either. So the general rule is that by the time of Coptic all glides were lost when directly following the stressed vowel and preceding a consonant. It is difficult to say whether the loss of all glides was contemporaneous or not. Some examples:
    wꜢḥ - to put ~ OⲨⲰϨS., OⲨOϨB.
    bjn - bad ~ ⲂⲰⲰⲚS., ⲂⲰⲚB.
    fꜢ.t - to carry ~ ϤⲒS.B.
  28. Exception: If in late compounds a stressed syllable loses stress secondarily, /j/ can move into an atonic syllable: jtrw-ꜥꜢ - big river > ⲈⲒⲈⲢOS., ⲈⲒⲀⲢO - Nile, a compound of 1) jtrw "river, Nile" > ⲈⲒOOⲢS., ⲒOⲢB. "canal" and 2) ꜥꜢ - big, great > -OS.B.
  29. Examples of this is given in Carsten Peust's book pg 149.
  30. Osing (1976a: 28-30) assumes that the development is conditioned by the quality of vowels which he assumes to have intervened between both glides. His rules can be summarized as follows: 1) if one of the glides was -w-, then it was preserved if the intervening vowel was -a-, 2) as an exception to this rule, the sequence -āwaw renders -OⲨⲒ(Ⲉ) ~ⲰⲒ if the intervening vowel was -i-, then j was preserved unless the stressed vowel of the word was i; 4) in all remaining cases (including the cases of intervening -it-), none of the consonants was preserved. Schenkel (1979) discusses Osing's rules in detail and proposes some modifications (Schenkel 1979: 389). However, evidence for determining these intermediate vowels is weak; according to our reconstruction there were no such vowels at all... ... Carsten Peust pg 146.
  31. Carsten Peust pg 150.
  32. Carsten Peust pg 146.
  33. Or in other positions sometimes in an unpredictable fashion
  34. Carsten Peust pg 120.
  35. Carsten Peust pg 120.
  36. Carsten Peust pg 168.
  37. An unexplained exception is šꜢs "nomad" > ϢⲰⲤ "herdsman".
  38. There is no assimilation in wsḫ "to be large ~ OⲨⲰϢⲤS.B. (with metathesis)
  39. Carsten Peust pg 169.
  40. That is if /w/ was pronounced as a (semi-)consonant not if in the hierpglyphs /w/ represented a vowel.
  41. It is better to analyze this as a feminine form collapsed and adopted as the masculine form rather than an assimilation due to contact with /w/.
  42. To date this theory does not fully explain the Coptic vowel H specifically in adjectival or 2 radical qualitative forms.
  43. A phonetic evolution which probably did not affect the phonological level is /i:/ > [e:] in proximity of /ʔ/ and /j/.
  44. It is possible but less likely that a long stressed u may have turned into a long stressed ʊ (at some point in the Middle Kingdom or Late Kingdom) and not long after that it may have been pronounced as an unrounded near-close near-back vowel (ɯ̽) due to the other vowels being unrounded. Though since a long stressed ʊ is rather rare among North African languages (ɯ̽ is even more rare) this is majorly an unproven hypothesis but just as possible.
  45. Some scholars, for example Helmut Satzinger believe accent was in free variation in the earlier phases of the Egyptian language: FOLIA ORIENTALIA vol 49 2012, On Egyptian Participles and Nomina Agentis by Helmut Satzinger pg 473.
  46. pg 19
  47. pgs 50-51
  48. pgs 167-168 James P. Allen, Middle Egyptian: An Introduction to the Language and Culture of Hieroglyphs 2nd Edition
  49. Loprieno, Antonio. A Linguistic Introduction pg 89
  50. Loprieno, Antonio. A Linguistic Introduction pg 79