Home ] Up ]


Suffixes originally consisted only of vowel: iod and kamatz for the first and the second person. The affix vowels are similar in nouns, past and future tense, and imperative mood. Consonants, on the contrary, are wildly different and allowed to drop out. That the same ת is the past tense suffix both in the first and the second person also suggests that consonants are non-essential.

Unaccented vowel suffixes would drop out, as did the case suffixes. Confirming consonants were added to prevent the reduction, catAvi – catav’, but catAvti.[1]


Conjugative suffix in 3ms verbs is unnecessary because it could be figured out by elimination.

Verbs usually have subject, and do not need unconjugated form. In the rare instances of absent subject, 3ms form was used, rained – it rained – he rained.

Nouns, however, often lack possession, and need unpossessed form, davar. To avoid confusion with unpossessed form, 3s nouns (unlike verbs), have suffix, u – hu – au – o.


Late 1p suffix is based on consonant nun rather than a vowel. The choice of nun was mistakenly modeled on the suffix of abstract plurality –on (helbon), where nun is only a confirming vowel. Suffix –on was changed into –no to conform to the verb suffix mold CV. Unaccented [o] became shuruk, -nu.


The original 2p suffix –ten became feminine through association of נ with nekiva.[2] The new masculine -tem form was based on מ from the –im suffix.


While pronounceable after the open syllable (catavtini), נ dropped out in 1s of nouns, dvarni - dvari. In 1p of nouns, נ was preserved by epenthesis, dvarenu.[3]


When possessive verb suffixes were decided upon, they had to be distinguished from conjugative: otherwise, 3ms verbs with 1s possessive –ti would be indistinguishable from 1s conjugative –ti. Non-essential suffix consonants were changed to נ and כ.

נ was added to 1s possessive suffix by analogy with 1p conjugative. The 2s כ was likely chosen as the next letter in alphabet after י of the 1s suffix.


Plural suffixes are compound, davar – d’varchA – d’v’rchaIm – d’v’rchEm (tzere) – d’varchem (epenthesis of patah at the expense of tzere).

How do we know that seghol in –chem appeared from tzere, not from schwa, ch’m - chem? Accented schwa becomes tzere, dvarEnu. Accented seghol only appears in place of long vowel reduced because of the epenthesis, cotev (tzere) – cotevet.

Plural suffixes combine the respective 2-3s suffixes with gender-specific plurality, cha-im – chem, ch-en – chen. 3p suffix hu-im (3ms+masc.pl.) was molded hem after the 2p suffix chem.


The first feminine suffix was ה in 3fs. In paal, 3fs is modeled on other stems: catva (patah shortens) instead of ctava (pro-pretonic tzere shortens) follows hitcatva and nicht’va.

2fs was modeled on 3fs, catavh (catva) – catavt.

Consonantal clusters are lost in speech, catavt – cata’t. The Masoretes suggested dagesh, a stop, to preserve final consonant, catavt – catav.t.  When the nearby vowel is long enough, it splits in epenthesis, cotevt – cotevet. In nouns, accent shift elongated the schwa, davarch – davar’ch – d’varEch.


The duplicity suffix –aim is a combination of the second- and first-person suffix’ vowels: -ca and –i, “your and my” referring to two. Diphthongs are accented on the first vowel sound, and -Ai requires a confirming consonant, מ.

The abstract plurality suffix –ot is a combination of the second- and third-person vowels: -ca and –hu, “you and he” referring to faraway, abstract many, -au – o – on (Helbon) – ot.

Constructus plural suffix derives from the duplicity suffix, ai – e. Open accented vowel was closed with iod to prevent post-tonic blurring, *divrE-torAh – divret-‘rah. Compare iod in Ashkenazic dOivor.

Constructus is pronounced on a single breath, and stress on the tzere suffix is relatively weak. Elsewhere, plural suffix tzere has to be closed firmer to avoid the post-tonic blurring, ai – e – em. The choice of mem was mistaken because of its stop, dvarem – dvare.m. When mem is pronounced distinctly, it draws accent from tzere and reduces it into hirek, dvare-m – dvarim.


Plural nouns follow predictable pattern, dvarei – dvareii – dvaraii (plural and 1s suffixes; two iod’s form heavy syllable, and reduce tzere to patah) – dvarai. Similarly, 2fp dvaraich derives from plural ei and 2fs caf suffixes.


New form of the future tense was built with prefixes א, ת, ה (modeled on 1-3ms pronouns) + imperative verb. Weak ה turned into י (compare rotzeh – raziti).

Confusingly, hey-derived iod denotes both feminine (2fs) and the third person (3ms) future tense. The FT inventors had to accept the ambiguous role of hey (iod) because the other characteristic component of 3ms suffix, waw, was already reserved for tense reversal.

Prefixes in the FT denote person, and suffixes – gender. By this logic, 3fs FT should be הct’vה (ict’vi). Two ה were added up into ת, creating apparent similarity to 2ms form, tictov.

2-3mp prefixes are the same as in 2-3ms. Shuruk as suffix of plurality follows the 3p past tense suffix.

2-3fp prefix follows 2-3fs. In the suffix –na, ה means feminine. נ is the same feminine suffix of plurality as in the 2fp past tense. Nun was taken for the suffix of plurality instead of shuruk (2-3mp) to avoid the uncomfortable uה suffix.


In fem. sing. present tense of verbs, accent on tzere made suffix unaccented, *cotEva – cotEv’. Tav was substituted for hey to prevent the reduction.


Pronouns were constructed by adding possessive suffixes to the noun אנ, living essence: ani is my an. Hatef proves the reduction induced by suffix. Similarly, אנתה and אנחנ׀ with interpolated ת and ח.[4]


Pronominal and case suffixes are related. Iod marks genitive case which conveys possession. In speech, the most common possessive relationship is first person, my thing, also marked by iod.

Kamatz marks accusative case which often applies to interlocutor’s, your things.

Shuruk marks nominative case, and also the most distant, often abstract his/its possession. Nominative is the basic case, and the third person is the basic verb form.


Just like affixes, prepositions have coherent meanings which seem to fluctuate only because of approximate translation. For example, l always denotes approaching and thus dative case, giving to. In bahar lanu, he chose us, l is seemingly accusative. The translation, however, is inaccurate. Bahar has a sense of thrusting in the direction, and is properly accompanied by directional or dative preposition l. Bahar lanu should be properly understood as, he made a divide for us.

[1] Tav could be chosen for its stop-like quality which firmly protected vowel, but also for the metaphysics of being the last letter in alphabet. Nun, another confirming consonant, is the middle letter of alphabet if we exclude vowelized aleph, waw, hey, and iod.

[2] The נ variation of 2-3p suffixes could be originally Aramaic version of Hebrew מ suffix.

[3] Compare ben – bant (feminine suffix tav, reduction of tzere in heavy syllable) – bat.

[4] Also, an - m’ain; iod closes semi-open (wean nun) accented vowel, as in dOivor.