The seldom discussed III-t (ל״ת) class of Biblical Hebrew Verbs, 2022


As a Ph.D. student at the University of Texas, we had a class called “Turbo Hebrew,” in which we read through the entire Hebrew Bible over the course of four semesters. Our primary homework each week was simply to read the text and note any peculiarities of grammar we came across in our reading. We usually had enough time for each student to bring up 2-3 questions that we would discuss as a class. Right up there with teaching Biblical Hebrew as a living language and learning Modern Hebrew, the countless hours spent in preparation for this class and the classroom discussions which followed were one of the most impactful and fruitful experiences for growing in my knowledge of Biblical Hebrew. There is no substitute for long and in-depth exposure to authentic language material.

Discovering the III-t (ל״ת) Class of Verbs

During this time, I remember coming across the following form in I Kings:

I Kgs. 1.15b

‏וַֽאֲבִישַׁג֙ הַשּׁ֣וּנַמִּ֔ית מְשָׁרַ֖ת אֶת־הַמֶּֽלֶךְ׃

‘And Abishag the Shunamite was attending to the king.’

The verb highlighted above is a feminine participle form of the piʿʿel/piʿʿal verb שֵׁרֵת-יְשָׁרֵת ‘to serve; to minister; to atttend’ from the root שׁר״ת. The masculine singular form of the participle is מְשָׁרֵת, just as in Modern Hebrew. Now as a Modern Hebrew speaker, I would have expected the feminine singular form of the verb to be מְשָׁרֶתֶת, just like the strong pattern of other piel participles in Modern Hebrew like מְדַבֵּר-מְדַבֶּרֶת ‘speaking’, מְבַקֵּשׁ-מְבַקֶּשֶׁת ‘asking’, and מְחַפֵּשׂ-מְחַפֶּשֶׂת ‘searching’. This is, in fact, exactly how the piel verb שֵׁרֵת conjugates in Modern Hebrew: i.e., מְשָׁרֵת-מְשָׁרֶתֶת-מְשָׁרְתִים-מְשָׁרְתוֹת.

So how exactly do we explain what is going on with the form מְשָׁרַת found in I Kings? Remarkably, there is only one other III-t (ל״ת) feminine singular participle in all of the Hebrew Bible:1

I Sam. 16.15b

‏הִנֵּה־נָ֧א רֽוּחַ־אֱלֹהִ֛ים רָעָ֖ה מְבַעִתֶּֽךָ׃

‘Look, now, a harmful spirit from God is tormenting you.’

This form gives us all the clues we need to understand what is going on with the III-t (ל״ת) forms in Biblical Hebrew. Indeed, the form מְבַעִתֶּךָ is a feminine singular participle of the piʿʿel/piʿʿal verb בִּעֵת-יְבַעֵת ‘to terrify; to assail’ from the root בע״ת. Similar to the example from I Kings, it constitutes a III-t (ל״ת) form of a feminine participle in the piʿʿel/piʿʿal binyan. There are, however, several important differences:

  • It has a 2MS suffix attached, which results in stress being on the following syllable.
  • The final vowel of the stem, which is thus unstressed, has a ḥireq vowel instead of a pataḥ.
  • The final /t/ of the stem is geminated (i.e., has dagesh).

Presumably, then, the participle מְשָׁרַת would exhibit similar behavior with a suffix attached: i.e., הִיא מְשָרַת אֹתוֹ ← הִיא מְשָׁרִתּוֹ ‘she is attending to him’. Similarly, the feminine participle of the verb בִּעֵת-יְבַעֵת would likely exhibit a form akin to מְשָׁרַת without a suffix: i.e., הִיא מְבַעִתֶּךָ ← הִיא מְבַעַת אֹתְךָ ‘she is tormenting you’.

In sum, then, the final vowel of the stem is /a/ before a single /t/ when stressed (מְשָׁרַת) and /i/ before a geminated /tt/ when unstressed (due to a following suffix) (מְשָׁרִתּוֹ).

The Parallel Form בַּת ‘daughter’ → בִּתּוֹ ‘his daughter’

Now this patterning of vowels, gemination, and stress parallels that of the noun בַּת ‘daughter’. In the singular form with no suffixes, the noun has a pataḥ vowel. However, when a suffix is added, stress is on the suffix, the vowel changes to ḥireq, and the /t/ is geminated: בַּת ‘daughter’ → בִּתּוֹ ‘his daughter’.

Historically, however, it is important to note that the word בַּת is ultimately derived from the form */bint/ with an original */i/ vowel. Eventually, the /n/ of the noun assimilated to the final /t/ of the noun so that the form became */bitt/ (or */bɛtt/). The development from */bint/ to בַּת, then, has often been attributed to the sound change known as Philippi’s Law, which in its simplest form states that */i/ shifts to */a/ in word-final stressed syllables that are doubly closed: i.e., *í → *á / _CC. Now, admittedly, this is a bit of an oversimplification of things. After all, Philippi’s Law has been reformulated countless times since its inception in 1878—the most recent (and probably best) reformulation of it is by Suchard in his recent 2020 work on the Biblical Hebrew vowels.2 But for our purposes here, we do not need to get into all the details.

What is important to note is that a form with an original */i/ vowel ending in a final geminated */tt/ would be expected to exhibit the same behavior as a word like בַּת, for which such patterning is widely attested.

The Historical Development of III-t (ל״ת) Participles

How, then, should we explain the historical development of III-t (ל״ת) participles in Biblical Hebrew? Well, we should first recognize that the final ת ֶ – ending that we are accustomed to see on feminine singular participles is relatively innovative. The original form of a feminine singular participle in the qal binyan, for example, would have been something like */qoːṭil-t/ as opposed to the masculine form */qoːṭil/. The seghol vowel between the final root consonant and the feminine /t/ ending is actually relatively new from a historical perspective. It does not appear to have been present during the earlier parts of the biblical period. Even as late as the Roman period in Palestine, we find that at least certain Biblical Hebrew traditions still had no epenthetic vowel there. Note that the word קֹהֶלֶת ‘preacher; teacher’, though not a feminine form, exhibits the same pattern and is transcribed as κωελθ in Greek.3

It was only at a later stage of the language that these final consonant clusters were broken up with an epenthetic (or “helping”) vowel. Such “helping” vowels are actually the reason why we say melech for מלך ‘king’ and not melk, which was the more historical form, as we can tell from ancient Greek transcriptions of Hebrew which render the word as μελχ.4

So the development of a participle like כּוֹתֶבֶת/כֹּתֶבֶת ‘writing (fs)’ would have proceeded as follows: */koːtib-t/ → */koːteb-t/ → */koːtebet/ → כֹּתֶבֶת. The final seghol before the feminine /t/ was introduced to facilitate the pronunciation of the final consonant cluster.

However, this same development could not have occurred with III-t (ל״ת) verbs. Indeed, in the biblical period, we should reconstruct the masculine participle משׁרת as */maʃarrit/ and the feminine participle as */maʃarrit-t/. Because the final root consonant was a /t/, it would have simply assimilated to the feminine /t/ ending and resulted in a geminated consonant. As such, a doubled consonant cannot be split up by an epenthetic. Rather, the form proceeded along the same development as בַּת (see above). This is due to the final gemination of the /tt/, which is still visible in forms like מְבַעִתֶּךָ (see above).

Should III-t (ל״ת) be a verb class across the binyanim?

Presumably, then, the same development would have affected other binyanim with the same pattern (i.e., a typical segholate ending for the feminine participle) as well. Even though we only have feminine participles of III-t (ל״ת) verbs attested in the piʿʿel/piʿʿal binyan, it is plausible that the same development would have affected other verbs as well.

Note, for example, that a verb like כָּרַת-יִכְרֹת ‘to cut’ would almost certainly have to exhibit the same development: i.e., */koːrit/ → כּוֹרֵת ‘cutting (ms)’ but */koːrit-t/ → כּוֹרַת ‘cutting (fs)’ but כּוֹרִתּוֹ ‘cutting (fs) it (ms)’. While we cannot be certain about such hypothetical forms, all the attested patterns would seem to suggest that such forms would be the most likely.

Evidence from the III-d form יוֹלַדְתְּ

There is, in fact, one final piece of evidence that may support this point. Note that in three cases in the Hebrew Bible, the feminine participle of the qal verb יָלַד-יֵלֵד ‘to birth’ is vocalised as יֹלַדְתְּ instead of the expected יֹלֶדֶת (cf. Gen. 17.19; Lev. 12.7; Isa. 7.14; Jer. 15.9; Jer. 31.8 for the typical form):

Gen. 16.11b

‏הִנָּ֥ךְ הָרָ֖ה וְיֹלַ֣דְתְּ בֵּ֑ן

‘Look, you are pregnant and will bear a son.’

Judg. 13.5a

‏כִּי֩ הִנָּ֨ךְ הָרָ֜ה וְיֹלַ֣דְתְּ בֵּ֗ן

‘For look, you are pregnant and will bear a son.’

Judg. 13.7

‏הִנָּ֥ךְ הָרָ֖ה וְיֹלַ֣דְתְּ בֵּ֑ן

‘Look, you are pregnant and will bear a son.’

While some scholars argue that יֹלַדְתְּ is a mixed form between the qaṭal יָלַדְתְּ and the participle יֹלֶדֶת (see, e.g., Jouon & Muraoka 2009, §89j), there may actually be a phonological explanation. Indeed, while a mixed form is theoretically possible, it seems unlikely to me. Appealing to something as being a “mixed” form is often just a way of explaining away difficult morphology—it is overused as an explanation in the reference grammars and should only be a last resort. Rather, I think that the following verse suggests that the participle would indeed be the expected form in such a context:

Isa. 7.14

‏הִנֵּ֣ה הָעַלְמָ֗ה הָרָה֙ וְיֹלֶ֣דֶת בֵּ֔ן

‘Look, the virgin/maiden will be pregnant and will bear a son.’

If a form like יֹלַדְתְּ is not a mixed form, then, but rather a participle, how do we explain the unusual ending? We may find a clue in an unusual form from the root יל״ד attested elsewhere in the Bible:

I Sam. 4.19

וְכַלָּת֣וֹ אֵֽשֶׁת־פִּינְחָס֮ הָרָ֣ה לָלַת֒

‘And his daughter-in-law, the wife of Phineas, was pregnant, about to give birth.’

Note that the regular infinitive construct form of the verb יָלַד-יֵלֵד is לֶדֶת. As such, we would expect the form here to be realized as לָלֶדֶת. The attested consonantal text ללת, however, appears to reflect assimilation of the final /d/ of the root to the following feminine /t/ of this particular infinitive pattern. Like our preceding examples, the final pataḥ vowel is presumably the result of the final syllable being doubly closed. Note the likely development with assimilation of */d/ to */t/ and the resulting final gemination: */la-lidt/ → */la-litt/ → */laː-litt/ → לָלַת.

It seems, then, that because */d/ and */t/ were pronounced with the same place of articulation and only differed in voicing, that assimilation was not uncommon. Note that this has parallels in Qurʾanic Arabic as well, where a verb like وُلِدتُّ wulidtu is actually pronounced as [wulittu] (Surat Maryam, Ayat 33) with assimilation of the /d/ of the root to the /t/ of the ending.

Therefore, if we return to the Biblical Hebrew participle יֹלַדְתְּ, we might suggest the following development which, after the assimilation of /d/ to /t/, would progress in the same way as III-t (ל״ת) participles like כֹּרֵת-כֹּרַת, מְשָׁרֵת-מְשָׁרַת, etc.: */joːlid-t/ → */joːlit-t/ → יֹלַת. However, because the consonantal text of the MT had a dalet written in the form as וילדת, the vocalization יֹלַת was imposed over the existing consonantal text and thus resulted in יֹלַדְתְּ. If this is the case, it would supply additional evidence that this phenomenon of III-t (ל״ת) being its own class also applies in the feminine participle of the qal binyan.

Conclusions: How should this affect my teaching?

If we assume that the patterns and sound rules involving III-t (ל״ת) verbs are consistent across the binyanim, then we should by all means make sure to keep to this morphology when teaching and speaking Biblical Hebrew.

Therefore, though largely unattested, our paradigm of III-t (ל״ת) verbs will differ from the strong verbs in the following instances (each exhibited with a relevant common III-t verb as an example):

FormMSFSFS + suffix
qal participleכֹּרֵתכֹּרַתכֹּרִתּוֹ
piʿʿel/piʿʿal participleמְשָׁרֵתמְשָׁרַתמְשָׁרִתּוֹ
nifʿal participleנִכְרָתנִכְרַת?
hifʿil participleמַשְׁחִיתמַשְׁחִיתָה or מַשְׁחַתמַשְׁחִתּוֹ
hitpaʿʿel/hitpaʿʿal participleמִתְעַשֵּׁתמִתְעַשַּׁתמִתְעַשִּׁתּוֹ

Accordingly, using forms like כֹּרֶתֶת and מְשָׁרֶתֶת should probably be regarded as importing Modern Hebrew morphology into Biblical Hebrew, though it is not difficult to understand how such forms could develop. Rather than constituting a continuation of the historical form, they reflect a reconceptualization of the feminine ending ת ֶ – added on to the base stem.

  1. This excludes a couple cases of a middle weak participle from the root מו״ת, which are not relevant here.
  2. For more on Philippi’s Law, see Benjamin D. Suchard, The Development of the Biblical Hebrew vowels: Including a Concise Historical Morphology (Leiden: Brill, 2020), 141–167.
  3. See Benjamin Kantor, τὸ ἑβραϊκόν TO HEBRAIKON: A Critical Edition of the Second Column (Secunda) of Origen’s Hexapla (Leuven: Peeters, forthcoming).
  4. See Benjamin Kantor, τὸ ἑβραϊκόν TO HEBRAIKON: A Critical Edition of the Second Column (Secunda) of Origen’s Hexapla (Leuven: Peeters, forthcoming).

Related Articles


Leave a Reply