The morphology of classical Latin

7. Verbal morphology

The Latin verb is conjugated, i.e. adds to verbal lexeme some endings, which, being different from the nominal or casual endings, are called verbal final endings or verb-terminations. Traditionally, grammars say that every form of Latin verbs is made up of two parts which they called respectively stem and ending. They assume that every verb has three stems: the present stem, which may be found by dropping -re in the active present infinitive, therefore amā-, monē-, lege-, cape-, and audī-, the perfect stem, as respectively amāv-, monu-, lēg-, cēp-, and audīv-, and the supine stem, as respectively amāt-, monit-, lect-, capt-, and audīt-. The so-called present stem of a verb corresponds to that which we will call the signifier of the verbal lexeme or lexical morpheme. The set of the verbal forms which a verb can take is called its conjugation.

Traditionally, Latin grammars describe the conjugation of Latin verb with some semantico-morphological categories, as voices, persons, moods, tenses, and aspects. These different traditional categories having each some formal specific features can therefore be considered as at least morphologically justified.

Grammars teach that there are three voices in Latin, the active, the passive and the deponent. But from a morphological point of view, the three voices amount to only two; because the so-said deponent verbs are verbs which have passive forms, but no active forms (grammars prefer to say they have a passive form, but an active meaning). Thus the lexeme “love”, in Latin amāre, has an active form amō “I love” fronting the passive form amor “I am loved”. But the verb “imitate”, in Latin imitārī, has a passive form imitor, which certainly means “I imitate”, but is formally the same as the passive amor. It is therefore called deponent, and not passive, because it seems to mean the same thing as the form active amō, but there is, in Latin, no active form *imito. And when by accident an active form exists beside the deponent form, as mereō beside mereor “merit”, opīnō beside opīnor “think”, ou cunctō beside cunctor “hesitate”, there never is the difference of meaning which contrasts amō with amor.

Each of the voices has two series of moods and tenses, which are called infectum (“unfinished”) and perfectum (“finished”) since antiquity and considered by the moderns as aspects. Each of the series has at least two and at the most three personal moods, called respectively indicative, subjunctive and imperative, the imperative having only two persons and existing only in the infectum, and at least two non-personal moods, called respectively infinitive and participle, the participle belonging to the infectum in the active, but to perfectum in the passive, at least theoretically. To the infectum infinitive are related two other moods, traditionally called gerund and supine.

Each mood has between one and three tenses, which are called present, imperfect and future in the infectum, and perfect, pluperfect and future perfect in the perfectum. It would be truer from a morphological point of view, if they were called respectively perfectum present, perfectum imperfect, and perfectum future. The indicative has the three tenses and the subjunctive only the first two. The imperative and the infinitive have in the passive infectum only one tense, the present; but in the active infectum they add the third tense, i.e. the future. The participle has two tenses in the active infectum, the present and the future, but one only in passive perfectum. The different traditional categories are combined according to the traditional system of the following table:

indicative subjunctive imperative participle infinitive
infectum present present present present present
imperfect imperfect
Future future future future
perfectum perfect perfect perfect
pluperfect pluperfect
future perfect

Traditional categories of the Latin verb:

In the traditional categories we should distinguish two kinds. There are the categories which don’t correspond to any morphological segment, as active, infectum, indicative, and present, and the categories which correspond to a morphological unit (i. e. a set of morphological segments in complementary distribution), as the Passive voice, the Perfectum aspect, the Subjunctive, Imperative, Infinitive and Participle moods, the Imperfect and Future tenses, and the six Persons. These morphological units, which we shall write with a capital initial from now on, in order to distinguish them from the traditional categories which show them, constitute classes or paradigms that we can call with traditional names. There is thus the paradigm of tenses, which is formed by the two morphological units of the Imperfect and Future, the paradigm of moods, which is constituted by the Subjunctive, the Imperative being only a variation of the Subjunctive, the paradigm of voices, represented by the only Passive, and the paradigm of persons, which is constituted by the Person 1 (= the first person of singular), Person 2 (= the second person of singular), Person 3 (= the third person of singular), Person 4 (= the first person of plural), Person 5 (= the second person of plural), and Person 6 (= the third person of plural).

All these morphological units are the signifier of a morpheme, except the Subjunctive which is the signifier of two homonymous morphemes, because it corresponds to two different signifieds, which can be defined by the semantic feature of “willpower” and “possibility”, the first morpheme leading to the variant ne of the negation, and the second, the variant nōn of the negation. To the Imperfect will be attributed the signified of “non-actual”1), to the Future the signified of “planned”, to Perfectum the signified of “accomplished”, and to the Passive the grammatical signified of “intransitive”. Finally to Person 1 will be attributed the signified “speaker”, to Person 2 the signified of “interlocutor”, to Person 3 the signified “nether speaker nor interlocutor”, to Person 4 the signified “speaker and others”, to Person 5 the signified of “interlocutor and others”, to Person 6 the signified “several people nether speaker nor interlocutor”.

As for the categories which have no morphological segment, they have neither signified nor signifier. They are not morphemes. Thus for instance the word “present” refer to the verbal forms which have neither the morpheme of “no actual” nor the morpheme of “planned”, and corresponds therefore to an absence of tense morpheme:

indicative Subjunctive Infinitive
present amā–tis am–ē-tis amā–re
Imperfect amā-bā-tis amā-rē-tis
Future amā-bi-tis

And so, the word indicative is used for all the verbal forms without morpheme of mood:

present Imperfect Future
indicative amā–tis amā–bā-tis amā–bi-tis
“possibility” am-ē-tis amā-rē-tis
“willpower” am-ē-tis amā-rē-tis

As for the word infectum, it is used for all verbal form which is characterized by the absence of the Perfectum aspect morpheme, which is clearly shown by the following examples:

present Imperfect Future Infinitive
infectum amā–tis amā–bā-tis amā–bi-tis amā–re
Perfectum amā-uis-tis amā-uer-ā-tis amā-uer-i-tis amā-uis-se

We can represent the combinatory of the different morphemes of the Latin verbal system with the following table, where only the personal verbal forms are represented, without taking account of the possible restrictions of distribution.

Voice Aspekt Moods Tenses Persons
/Pers 1/
/Pers 2/





/Pers 3/
“speaker +”
/Pers 4/
“interlocutor +”
/Pers 5/
“he +”
/Pers 6/

Combinatory of the verbal morphemes

This combinatory corresponds to the following generative rules:

  • 1. Endings of persons

In the singular, the morphological unit of first Person corresponds to the morphological segment in the indicative present of all the verbs except esse “be”:

  • am-ō, “I love”; mone-ō, “I warn”; leg-ō, “I read”; capi-ō, “I take”; audi-ō, “I hear”


  • su-m, “I am”; ab-su-m, “I am away”; pos-su-m, ”I can”,

and in the future in -b- or :

  • amā-b-ō, “I shall love”; monē-b-ō, “I shall warn”; er–ō, “I shall be”; amā-uer–ō, “I shall have loved“; mon-uer–ō, “I shall have warned” .

In indicative perfect, a segment morphological is used:

  • amā-u-ī, “I loved, have loved”; mon-u-ī, “I warned, have warned”; lēg-ī, “I read, have read”; etc.

Everywhere else it is the morphological segment -m. The verb esse and its compounds therefore use the unmarked allomorph of the morphological unit of the first Person. All that corresponds to the following complex rule:

The second Person shows the morphological segment -s, with the exception of the imperative, where a segment appears:

  • amā, “love thou”; monē, “warn thou”; lege “read thou”; etc.

and the perfect indicative, which shows an allomorph –tī:

  • amā-uis-tī, “you loved, had loved”; lēg-is-tī “you read, had read”; etc.

The morphological unit of the third Person corresponds only to the morph -t. And so, on the plural, the morphological unit of fourth Person always corresponds to the morphological segment -mus, and the 5th Person always to the morphological segment -tis, with the exception of the imperative, which uses the allomorph -te:

  • amā–te, “love ye”; monē–te, “warn ye”; legi–te “read ye“; etc.

The unity of the 6th Person corresponds to the morphological segment -nt, which shows a variant -unt after a consonant or closed vowel:

  • leg-unt, “they read”; capi-unt, “they take”; audi-unt, “they hear” (phonetic realization of /audi:-unt/); flu-unt “they flow”; er-unt, “they will be” (phonetic realization of /es–unt/); amā-b-unt “they will love”; amā-uer-unt, “they loved, have loved” (phonetic realization of /ama:-uis-unt/).

And, beside amāuerunt, there is a variant amāuēre, which shows an allomorph -ēre of the 6th Person, being the phonetic realization of /ama:-u-e:re/).

  • 7.2. The infectum

The infectum is defined by the absence of aspectual morpheme, which is particularly visible with the minimal oppositions amā–tis ~ amā-uis-tis, ama–t ~ amā-ui-t, and also amā–ba-m ~ amā-uer-a-m, amā–re-m ~ amā-uis-se-m, when we know that -ba- and -a- are allomorphs of the morpheme Imperfect, and [r] an allophone of /s/. That means that amā-tis and ama-t are constituted by only two morphemes: the lexical morpheme and the morpheme of Person.

  • 7.2.1. The different types of infectum

Traditionally, grammars base the identification of four or five Latin conjugations on the stems of the infectum. Varro was the first one who seemed to distinguish three conjugations based on the vowel of the second person of the present:

  • Varro, L.L. 9,109: “meō meās; neō nēs; ruō ruis” .

Priscianus2) will speak of four conjugations, putting cupio cupǐs, curro currǐs in the third conjugation, and admitting munio munīs, esurio esurīs to a fourth conjugation. It is almost the modern system, which assumes four conjugations, but divides into two the third conjugation, speaking of a third mixed conjugation, because certain of its forms are like the fourth conjugation. Thus, we should have the 1st conjugation or conjugation in ā (type amā-re “love”), the 2nd conjugation or conjugation in ē (type monē-re “to warn” or delē-re “blot out” or nē-re “spin”), the third conjugation or conjugation called consonantal or thematic, because beside a historically consonantal stem it shows an allomorph which adds to this stem a vowel called thematic of tone e or i (type leg-ō, legi-s, lege-re “read”, phonetic realization of /legi-se/), the fourth conjugation or conjugation in ī (type audī-re “hear”), and a fifth conjugation which is called third mixed conjugation, because its stem always ends with an i like in the fourth conjugation, but an i which is always short like in the third conjugation (type capĭ-ō, capǐs, cape-re, phonetic realization of /kapi-se/).

  • 7.2.2. The different conjugations

There are certainly, from the morphological point of view, five types of verbs; but that doesn’t mean five sorts of conjugations. Actually, if we look at the morphological segments which combine with these verbs, we note that beside the morphological units which are represented by the same segments in all the verbs, there are others as the Imperfect, the Future, the present Participle, and the Person 6, which are represented each by two different morphological segments in the same way. The verbal lexemes which end or may end with an i, namely

  • leg-ō, legi-s, whose stem is legi- before a consonant, and leg- before a vowel,
  • capĭ-ō, capǐs, whose stem always ends with an ǐ,
  • audi-ō, audī-s, whose stem ends with an ī, [awdijo:] being the phonetic realization of /audi:-o:/

are really combined with the same morphological segments of Imperfect -ēbā-, Future /a: ~ e:/, Participle -ent- and Person 6 -unt. leg-ēba-m, leg-ēbā-s, etc., capi-ēba-m, capi-ēbā-s, etc., audī-ēba-m, audī-ēbā-s, etc.; leg-a-m, leg-ē-s, etc., capi-a-m, capi-ē-s, etc., audi-a-m, audi-ē-s, etc.; leg-en-s, leg-ent-is, capi-en-s, capi-ent-is, audi-en-s, audi-ent-is; leg-unt, capi-unt, audi-unt.

Instead of these morphological segments, the lexemes in ā- and ē- are combined with other identical segments: /ba/ for the Imperfect, /b ~ bi/ for the Future, -nt- for the Participle, and -nt for Person 6:

  • amā-ba-m, amā-bā-s, etc., monē-ba-m, monē-bā-s, etc.; amā-bō, amā-bi-s, amā-b-unt, monē-b-ō, monē-bi-s, monē-b-unt; ama-n-s, ama-nt-is, mone-n-s, mone-nt-is; ama-nt, mone-nt.

There are undisputably two conjugations, but, may be, three, because there are two differences between the conjugation of amā-re and monē-re. First, the verbs as amā- show a variant without the final a-, before a vowel, hence the verbal forms am-ō, “I love” and am-e-m, am-ē-s , etc. in the Subjunctive, while the verbs as monē-re always keep their final vowel , which is phonetically realized [e:] or [e] in accordance with the Latin phonological system:

  • moně-ō, monē-s, moně-t etc. in the indicative, moně-a-m, moně-ā-s, etc. in the Subjunctive.

But the variation /ama: ~ am/ is easily understandable, when we know that the vocalic clusters /ao/ and /ae/ do not exist in Latin; only the diphthongs /au/ and /ae/ exist in Latin. Another difference between the verbs in ā- and in ē-, the Subjunctive which corresponds to a morphological segment -ā-, for both conjugations:

  • leg-a-m, leg-ā-s, capi-a-m, capi-ā-s, audi-a-m, audi-ā-s, and mone-a-m, mone-ā-s, is represented by an allomorph -ē-, for the only verbs like amā-re:
  • am-e-m, am-ē-s, am-e-t, am-ē-mus, am-ē-tis, am-e-nt.

But if the Subjunctive was then the segment -ā-, it would be almost impossible to be distinguished with the indicative, the Person 1 excepted:

  • am ō, amā-s, ama-t, amā-mus, amā-tis, ama-nt.

Is that difference sufficient to admit a third conjugation? It seems that Varro thought so. It is also the point of view of Donatus, who taught that “there are three conjugations: the 1st one which has an a long before the final letter of the second person of the active present indicative and before the final syllable of the second person of the passive present indicative, as amo amas, amor amaris; the second one which has an e long before the final letter of the second person of the active present indicative and before the final syllable of the second person of the passive present indicative, as doceo doces, doceor doceris; and the third one, which has an i sometimes short sometimes long before the final letter of the second person of the active present indicative and, instead of the i, an e short or an i long before the final syllable of the second person of the passive present indicative, as lego legis, legor legeris, audio audis, audior audiris3)” .

It is probably a bad but traditional cutting up of verbal forms that leads to distinguish the conjugation in -ās from the conjugation in -ēs, whereas it is the same conjugation in -bam, -bō, -nt. But it is really interesting that Varo and Donatus do not distinguish between lego legis and audio audīs. That shows that unlike our Latin grammars, the famous thematic conjugation does not hold an aside position. It is not the consonantal conjugation any more, the verb leg- of which would schow a variant legi- thanks to a thematization morphological rule like:

It has become the conjugation of a verb in i legi- which may show a variant without i, as am- is an allomorph of amā-. So, in conclusion the Latin verbal system only has two infectum conjugations; the 1st is constituted with verbs ending with a not closed (open) vowel ā- as well as ă- (cf. dă-re) or ē-, the 2nd with verbs ending with a closed vowel ǐ- as well as ī-. Nevertheless there are in Latin four kinds of verbs: verbs in ā and verbs in ē, which follow the 1st conjugation, and verbs in ī and the verbs in ǐ, wich follow the 2nd conjugation, and amongst these verbs in i, there are verbs which always have this ǐ as capere, capi-ō and verbs which sometimes lose it, as dice-re dīc-ō, which takes again Priscianus’s classification, which put both verbs “cupio cupǐs, curro currǐs” or “legor legĕris, sequor sequĕris” in the same conjugation. And these both conjugations are accompanied by an additional conjugation without the final vowel, i.e. the conjugation of am-ō, amā-re, and the conjugation of leg-ō, lege-re (phonetic realization of /legi-se/). Why do such verbs as minu-ō, argu-ō, or linqu-ō belong to the same conjugation as leg-ō? As for linqu-ō, it is perfectly normal, since this verb ends with a consonant /kw/. But as for the other verbs, it is the same thing: the phonemes /u/ and /u:/ being phonetically realized [uw] before a vowel as /ru-o:/, /rui-s/ [ruwo:] “fall” and /minu:-o:/ [minuwo:] “reduce”.

The two infectum conjugations therefore correspond to the conjugation of monē-re and the conjugation of audī-re or cape-re, depending on whether the verbal lexeme ends with ī- or with ǐ-.

Table of the 1st conjugation:

indicative Subjunctive Imperative Participle Infinitive
…….. -tis

…….. -ā-s
…….. -a-t
…….. -ā-mus
…….. … -tis
…….. -a-nt


mone-n -s
…….. -nt-is


……. -bā-s
……. -ba-t
……. -bā-mus
……. … -tis
……. -ba-nt

……. -rē-s
……. -re-t
……. -rē-mus
…….. … -tis
……. -re-nt
monē-b -ō
……. -bi-s
…….. … -t
…….. … -mus
…….. … -tis
…….. -b-unt


…….. -tō-te
(-a, -um)

(-am, -um)

Table of the 2nd conjugation:

indicative Subjunctive Imperative Participle Infinitive
… -tis

… -ā-s
… -a-t
… -ā-mus
… … -tis
… -a-nt


… -te
audi- en -s
… -ent-is


……. -ēbā-s
……. -ēba-t
……. -ēbā-mus
……. … -tis
……. -ēba-nt

……. -rē-s
……. -re-t
……. -rē-mus
…….. … -tis
……. -re-nt
… -ē-s
… -e-t
… -ē-mus
… … -tis
… -e-nt


… -tō-te
(-a, -um)

(-am, -um)

indicative Subjunctive Imperative Participle Infinitive
capi -s
…. -t
…. -mus
… -tis
…. -unt

… -ā-s
… -a-t
… -ā-mus
… … -tis
… -a-nt


… -te
capi- en -s
… -ent-is


……. -ēbā-s
……. -ēba-t
……. -ēbā-mus
……. … -tis
……. -ēba-nt

……. -rē-s
……. -re-t
……. -rē-mus
…….. … -tis
……. -re-nt
… -ē-s
… -e-t
… -ē-mus
… … -tis
… -e-nt


… -tō-te
(-a, -um)

(-am, -um)

But both these conjugations have a kind of variant without the final vowel of the lexical morpheme: it is, for the 1st conjugation, the lexemes in ā which lose this ā before a vowel (hence am-ō, am-e-m, am-ē-s), and for the 2nd conjugation, some lexemes in ǐ- which lose this ǐ- before vowel (hence leg-ō, leg-ēba-m, leg-a-m, leg-ē-s, leg-ā-s).

Variant without a-: the conjugation 1a:

indicative Subjunctive Imperative Participle Infinitive
…….. -tis

…….. -ē-s
…….. -e-t
…….. -ē-mus
…….. … -tis
…….. -e-nt


ama-n -s
…….. -nt-is


……. -bā-s
……. -ba-t
……. -bā-mus
……. … -tis
……. -ba-nt

……. -rē-s
……. -re-t
……. -rē-mus
…….. … -tis
……. -re-nt
amā-b -ō
……. -bi-s
…….. … -t
…….. … -mus
…….. … -tis
…….. -b-unt


…….. -tō-te
(-a, -um)

(-am, -um)

Variant without i- : the conjugation 2a:

indicative Subjunctive Imperative Participle Infinitive
leg -ō
…….. -tis

…….. -ā-s
…….. -a-t
…….. -ā-mus
…….. … -tis
…….. -a-nt


leg- en -s
…….. -ent-is


……. -ēbā-s
……. -ēba-t
……. -ēbā-mus
……. … -tis
……. -ēba-nt

……. -rē-s
……. -re-t
……. -rē-mus
…….. … -tis
……. -re-nt
leg -a-m
…….. … e-t
……. ē-mus
……. … -tis
……. -e-nt


…….. -tō-te
(-a, -um)

(-am, -um)

  • 7.3. The Perfectum

The Perfectum adds to verbal lexemes a morphological unit which is constituted with two morphological segments, which can be either contiguous or not. When they are contiguous, they fit in between the Verb and the Person, when they are not, the second segment only fits in between the Verb and the Person, while the first is in front of the Verb or modifies the Verb itself. The first segment corresponds to what is called the “perfect stem” by grammars: amā-u-, de-d- or ue-:-n-, which only appears in the first, third and fourth person of the perfect indicative:

  • amā-u-ī, amā-ui-t perfect of amā-re; de-d-ī, de-d-i-t perfect of dare “give”, uēn-ī, uēn-i-t perfect of uenī-re “come”, etc.

and the second segment is a sequence -is-, which is common to all the Perfectum conjugations

  • amā-uis-tī, de-d-is-tī, uēn-is-tī, etc.

There are five kinds of Perfectum and therefore five Perfectum conjugations.

  • 7.3.1. Perfectum in /uis ~ u/

The most productive type of Perfectum is the Perfectum in /uis ~ u/:

  • amā-uis-tī, amā-u-ī; delē-uis-tī, delē-u-ī; audī-uis-tī, audī-u-ī¸ etc.

But the morphological combination is realized [wer] before a vowel, because of the rhotacismus:

  • amā-uer-unt, “they loved”, phonetic realization of /ama:-uis-unt/ with the allomorph /unt/ after the consonant of Person 6, amā-uer-a-nt, “they had loved”, the phonetic realization of /ama:-uis-a:-nt/, amā-uer-i-nt, “they will have loved”, phonetic realization of /ama:-uis-i-nt/ and its first segment without the second segment shows a variant /ui/ before consonant:
  • amā-ui-mus and amā-ui-t.

The morphological combination /uis/ sometimes shows a variant /s/ without /ui/, which grammars wrongly call a “contracted form”, as no-s-tī “you know” beside nō-(ui)s-tī or amā-r-unt beside amā-uer-unt, amā-s-se beside amā-uis-se, consuē-r-a-t beside consuē-uer-a-t, flē-s-tis beside flē-uis-tis, etc. The same type of variation happens to perfects in -sis- in archaic and colloquial usage, as dixtī for ixis-tī, trāxe for trāxis-se, uixet for uixis-se-t, where it cannot be a question, of course, of a contraction. The morphological variant /s/ without /ui/ must be distinguished from the phonological rule which changes /u/ into after a stressed ī and before another i, and leads a contraction of both closed phonemes, when they are not separated by a morpheme border. This phonological rule gives audi-ī beside audī-u-ī, but not *audī, or audi-er-a-m beside audī-uer-am, but dīt-is and not *dīit-is, beside dīuit-is, “rich-(Gen.)”. But poets sometimes neglect the morpheme border which is an obstacle to the contraction and use forms as petīt, without shortening, instead of petiit, the usual perfect of the verb pet-ō, pete-re, “seek to obtain”.

  • 7.3.2. Perfectum in /sis ~ s/

The second type of Perfectum, which is traditionally called “perfectum sigmatic”, appears after a consonantal variant of verbal lexemes, particularly but not exclusively verbs such as /di:ki- ~ di:k-/ or /fingi- ~ fing-/:

  • dixis-tī, dix-ī, “I said, have said”; finxis-tī, finx- ī “I shaped, have shaped”
  • man-sis-tī, man-s-ī, perfect of mane-ō, manē-re, “I remained, have remained” with of course the phonetic realization [ser], when there is the rhotacismus: dixer-unt, “they said”, phonetic realization of /dik-sis-unt/, dixer-a-nt “they had said”, phonetic realization of /dik-sis-a:-nt/, etc. man-ser-unt “they remained, have remained”, man-ser-a-nt “they had remained”, etc. and the allomorph /si/ before a consonant:
  • dixi-t, “he said”; dixi-mus “we said”,
  • man-si-t, “he remained”; man-si-mus, “we remained”.

The two segments of the other Perfectum morphological units are not contiguous; the second segment -is- always appears between the verb and the Person morpheme. So for the first segment, it appears either inside the verbal stem in the form of a lengthening of the stem vowel, or before the stem in the form of a reduplication of its initial consonant

  • 7.3.3. Perfectum with a vowel lengthening

In this Perfectum stem, the vowel is lengthened, and the lengthening forms with the segment -is- the signifier of Perfectum morpheme, which therefore is what modern linguistics calls a discontinuous morpheme; we shall write it /: … is ~ : /:

  • /ue-:-n-is-ti:/ uēn-is-tī, and uēn-ī, perfect of the verb uenī-re, “I came, have come” with of course the phonetic realization [er], when there is the rhotacismus: uēn-er-unt, phonetic realization of /ue-:-n-is-unt/, uēneram, phonetic realization of /ue-:-n-is-a:-m/, etc. and the allomorph /: … i/ before consonant
  • /ue-:-n-i-t/, uēn-i-t, “he came, has come”, uēn-i-mus.

This lengthening changes the sound of vowels when it is applied on a stem in a. Thus the perfect of the verbs as ag-ō, age-re, “drive” or faci-ō, face-re, “make” is ēg-ī or fēc-ī.

  • 7.3.4. Perfectum with reduplication

The Latin grammars call perfect with reduplication the perfect forms of the verb cad-o, cade-re “fall” ,

  • ce-cid-ī, ce-cid-i-t, ce-cid-is-tī, ce-cid-er-unt (with the morphological alternation /kad ~ kid/) or morde-ō, “bite”; curr-ō, curri-s, “run”; disc-ō, disci-s, “learn”:
  • mo-mord-ī, mo-mord-i-t, mo-mord-is-tī, mo-mord-er-unt
  • cu-curr-ī, cu-curr-i-t, cu-curr-is-tī, cu-curr-er-unt
  • di-dicī di-dic-i-t, di-dic-is-tī, di-dic-er-unt (with the morphological alternation /disk ~ dik/ of the stem).

The reduplication consists in again copying the first consonant of the verb generally with an e vowel, sometimes with the same vowel as the verb: fe-fell-is-tī, fe-fell-er-unt from the verb fall-ō, falli-s, “deceive”, pe-pend-is-tī, pe-pend-er-unt from the verb pende-ō, “hang”, pe-pul-is-tī, pe-pul-er-unt from the verbe pell-ō, pelli-s, “push”; de-d-is-tī, de-d-er-unt from the verbe da-re “give”; tu-tund-is-tī, tu-tund-er-unt of th verb tund-ō, tundi-s, “beat”, to-tond-is-tī , to-tond-er-unt from the verb tonde-ō “cut”. When the initial is a consonantal group with sibilant sp or st, the verb stem loses its sibilant:

  • spo-pond-ī, “I gave a pledge”, and not *spo-spond-ī, from the verb spondeō;
  • ste-t-ī, “I stood”, and not *ste-st-ī, from the verb stā-re.

The reduplication shows a variant , when because of a composition it cannot be in the word initial position: re–spond-ī, “I replied”, perfect of the verb responde-ō, morphologically composed of sponde-ō, whose the perfect is spo-pond-ī, oc–cid-ī “I am done for” from the verb oc-cid-ō, oc-cide-re, morphemically composed of cad-ō, cade-re, whose the perfect is ce-cid-ī.

  • 7.3.5. Perfectum in / … is ~ /

When the first segment is erased and the simple verb does not exist any more, then the morpheme of Perfectum is formally represented only by the segment -is- or -er- in percul-is-tī, percul-is-tis¸ and per-cul-er-unt, percul-er-a-m, and by the segment -i- in percul-i-t, percul-i-mus, but indirectly by the variant ī of Person 1 percul-ī, “I struck”, since this variant only appears in the context of the Perfectum morpheme. Obviousely, percell-ō, “strike” is a compound verb with the prefix per-, like defend-ō, “protect” with the prefix de-, which can justify the deletion of the reduplication; but the simple verbs *cell-ō and *fend-ō do not exist in Latin, or do not exist any more. It is therefore necessary to assume straightaway that the Perfectum is simply in / … is/. The same is true for the non-compound verbs such as scand-i-t, “he climbed” (S. Sev., sanc. Mart. 14,2), where the forms with prefix a-scend-ī, de-scend-ī led to the loss of the reduplication for the simple, or tul-ī, the perfect of fer-ō, “carry”, which a real perfect with the reduplication te-tul-ī is found beside (cf. Plaut, Amph. 716, Cist. 650).

  • 7.3.6. Perfectum conjugations and infectum conjugations

There are therefore four kinds of Perfectum conjugations which are independent from the two or four infectum conjugations. It is impossible to know the Perfectum conjugation of a verb whose the infectum conjugation we know, and conversely. We can only say that all the verbs in ā-re (therefore from the 1a. conjugation) have a /uis/ Perfectum, except da-re, “give” and stā-re, “stand”, which have a Perfectum with reduplication (de-d-ī and ste-t-ī), and that the same morphological segment preceded by an /u/ in the stem, as in monŭī [monu-w-i:], was particularly related to the 1st conjugation, but could concern some verbs of the 1a, as uetŭī perfect of uetā-re, “forbid”, the 2a, as colŭī perfect of col-ō, cole-re, “live in”, as well as the 2nd, as aperŭī, perfect of aperī-re, “open”, or rapŭī, perfect of rapi-ō, rape-re, “snatch away”.

As for the other Perfectum¸ they concern only the consonantal stem, and therefore the verb of the 2a conjugation. But some verbs of the 1a, 1st and 2nd can have any allomorphs with consonantal final in the Perfectum, thus the Perfectums with reduplication ste-t-ī from stā-re, mo-mord-ī from mordē-re, pe-per-ī from pari-ō, pare-re, “give birth”, the sigmatics man-s-ī from manē-re, “remain”, inspex-ī from inspici-ō, inspice-re, “inspect”; sen-s-ī from sentī-re, “feel”, with lengthening uīd-ī from uīdē-re, “see”; uēn-ī from uěnī-re, “come”, fūg-ī from fŭgi-ō, fuge-re, “run away”. Thus, in order to conjugate a Latin verb, it is necessary to know what is called by grammars the Principal Parts of the verb, which show the different conjugations which it comes under. These Principal Parts are:

1.the present indicative in the first and second person (as am-ō, ā-s)

2.the present infinitive (as amā-re), which show its infectum conjugation (1a) – the infinitive being not really useful, but able to replace the first and second person for the 1st and 1a conjugations and for the verbs in ī-re of the 2nd

3.the perfect indicative (as amā-u-ī), which shows its Perfectum conjugation (/uis ~ u/)

4.the neuter of the perfect participle (as amā-t-um) or if it is not in use, the future active participle (as amā-tūr-us), wich shows its passive Perfectum conjugation.

  • 7.4. Tenses and personal moods

Like the infectum aspect, the indicative mood and the present tense are not some morphemes; they are even not a morpheme, since they do not correspond to a morphological unit. They simply are names on empty bottles. The present indicative is therefore morphologically constituted only by the verbal stem and a personal ending and morphemically by the lexical morpheme and a Person morpheme: amā-s “you love” = “love-you”, monē-s “you warn” = “warn-you”, legi-s « you read = « read-you ». The verbal stem shows then all the modifications involved by the Latin phonological system, as the neutralization of the quantity opposition in a long syllable closed by another else than the phoneme /s/: ama-t “he loves”, mone-t “he warns”, audi-t “he hears”, phonetic realizations of /amā-t/, /monē-t/ and /audī-t/ or before a vowel:

  • mone-ō “I warn”, phonetic realization of /monē-ō/.

To that is added the rule probably more phonological than morphological which erase the final ā of the verbal stem before a vowel. These different phonological rules explain the surface disparity between on the one hand am-ō and on the other hand ama-t, ama-nt and amā-s, amā-mus, amā-tis.

  • 7.4.1. The imperfect insert a morphological segment /ba:/ between the verbal stem and the personal ending, in the 1st conjugation, but a segment /e:ba:/ in the 2nd conjugation:
  • amā-bā-s, monē-bā-s, amā-bā-mus, monē-bā-mus, amā-bā-tis, monē-bā-tis
  • leg-ēbā-s, capi-ēbā-s, audi-ēbā-s, leg-ēbā-mus, capi-ēbā-mus, etc.

and an allomorph /a/ in the verb es-se and its compounds: er-ā-s “you was”, phonetic realization after rhotacismus of /es-a:-s/, pot-er-ā-mus “we could”, phonetic realization of /pot-es-a:-mus/. All these allomorphs, of course, show a short allophone before a word-final consonant else than /s/:

  • amā-ba-m, amā-ba-t, amā-ba-nt, monē-ba-m, monēba-t, monē-ba-nt
  • leg-ēba-m, leg-ēba-t, leg-ēba-nt, capi-ēba-m, capi-ēba-t, capi-ēba-nt, audi-ēba-m, audi-ēba-t, audi-ēba-nt
  • er-a-m, er-a-t, er-a-nt, pot-er-a-m, pot-er-a-t, pot-er-a-nt, etc.

But the Latins must have been tempted to bring together the verbs with a vowel long in the 1th conjugation, which would explain that beside audi-ēba-m a form like audī-ba-m, more rare than audi-ēba-m, is always found in Latin, but particularly in the archaic playwrights and poets of the Augustan age.

  • 7.4.2. As for the future, it inserts a segment /b ~ bi/ between the verbal item of the 1st conjugation and the personal ending, /b/ before a vowel and /bi/ before a consonant:
  • amā-b-ō, amā-b-unt, amā-bi-s, amā-bi-t, amā-bi-mis, amā-bi-tis

or a segment /ē ~ ā/ after a verbal item of the 2nd conjugation, ā in the first person, and ē in the others:

  • leg-ē-s, leg-ē-mus, leg-ē-tis

with the phonological shortening before a word-final consonant other than /s/:

  • leg-a-m, leg-e-t, leg-e-nt.

As for the verb es-se, it has a future -b- ~ -bi- without the b, therefore a future in / ~ i/, which is parallel to its imperfect -bā- without the b, i. e. only in -ā- : er-ō “I shall be”, phonetic realization of /es–o:/, er-i-s “you will be”, phonetic realization of /es-i-s/, etc. Like for the imperfect audī-ba-m, a future in -b-o was analogically created in the verb in ī-:

  • A. Ernout, 1953 Morph. hist., p. 162: in the archaic age: aggredī-b-or, amicī-b-or, aperī-b-ō, audī-b-ō,…, dormī-b-ō, …, mentī-bi-tur, nescī-b-ō, oboedī-b-ō, … , seruī-b-ō, …, subuenī-b-ō, uenī-b-ō, In the Ciceronian and Augustan age: ēsurī-b-ō, impertī-bi-s, inuenī-bi-t, lenī-b-unt, mollī-bi-t, peruenī-b-unt, uenī-b-ō.
  • 7.4.3.Subjunctive

In the infectum subjunctive, the present doesn’t use any formal segment as in the infectum indicative. Therefore in the present subjunctive, there is only a Subjunctive morphological segment, which is an -ā- in the 1st and 2nd conjugation, the 1a excepted; in order to be different from the verbal stem, the Subjunctive corresponds then to an -ē- morphological segment:

  • mone-ā-s, mone-ā-mus, mone-ā-tis, leg-ā-s, leg-ā-mus, capi-ā-s, capi-ā-mus, audi-ā-s, audi-ā-mus, etc.
  • am-ē-s, am-ē-mus, am-ē-tis

with, of course, a short allophone before a word-final consonant other than /s/:

  • mone-a-m, mone-a-t, mone-a-nt, leg-a-m, capi-a-m, audi-a-m, etc. am-e-m, am-e-t, am-e-nt.

The verb es-se and its compounds, and the verb uel-le “want” (phonetic realization of /uel-se/) use an -ī- subjunctive allomorph, or its phonological variant -ĭ-:

  • s-i-m, s-ī-s, s-i-t,… pos-s-i-m, pos-s-ī-s, pos-s-i-t, … uel-i-m, uel-ī-s, uel-i-t, nōl-i-m, nōl-ī-s, nōl-i-t (phonetic realization of /nō(n ue)l-ī-m/), etc.
  • 7.4.4. The imperfect subjunctive

The two morphological units correspond, in all the conjugations, to only one segment /se:/, which is therefore a morphological combination. According to the Latin phonological system, his segment is normally realized [se:] or [se] after a consonant:

  • es-se-m, es-sē-s, es-se-t, es-sē-mus, etc.

but [re:] or [re], or [le:] or [le] after a liquid (either r or l):

  • fer-rē-s (phonetic realization of /fer-se:-s/), fer-rē-mus, fer-re-m, fer-re-t, etc.
  • uel-lē-s (phonetic realization of /uel-se :-s/), uel-le-m, etc.

and [re:] or [re] after a vowel, by the rhotacismus:

  • amā-rē-s, monē-rē-s, lege-rē-s, cape-rē-s, audī-rē-s (phonetic realizations of /ama:-se:-s/, /mone:-se:-s/, /legi-se:-s/, capi-se:-s/ et /audi:-se:-s/), amā-re-m, monē-re-m, cape-re-m, audī-re-m, etc.
  • 7.4.5. The imperative, which in Latin exist only in the second persons, has, in the present, no formal segment and therefore corresponds to a morpheme. But, in the future, it shows a morphological segment -tō-. Besides it uses specific allomorphs of Person 2 and Person 5: a morpheme for Person 2 and a variant -te for Person 5:
  • amā-Ø “love thou”, amā-te “love ye”, monē-, monē-te, lege- (phonetic realization of /legi-/), legi-te, cape- (phonetic realization of /capi-/, capi-te, audī-, audī-te
  • amā-tō “thou shalt love”, amā-tō-te “ye shall love”, monē-tō, monē-tō-te, legi-tō, legi-tō-te, capi-tō, capi-tō-te, audī-tō, audī-tō-te.

Some lexemes in ĭ of the 2nd infectum conjugation lose this vowel in the imperative second person, which gives dīc, dūc and fac. The full forms are found in the colloquial discourse dīce (Plaut., Rud. 124), dūce (Plaut., Trin. 384), face (Plaut., Pseud. 18), but are usual in the compound verbs, as addice, addūce, confice, calface, even if Cicero sometimes uses a short form as ēdūc (Cic., Catil. 1,5).

The future imperative has also a third and sixth person, respectively

  • amā-tō “he shall love” and amā-ntō “they shall love”, legi-tō, leg-untō, capi-tō, capi-untō, audī-tō, audi-untō.

The future imperative was frequent in early Latin:

  • Plaut., Merc. 770: Cras petito; dabitur. Nunc abi, “Ask tomorrow, it will be given. Now, away with you!”

and in poetry and legal texts

  • Verg., egl. 3,76: cum faciam uitula pro frugibus, ipse uenito, “when I shall sacrifice a heifer for the harvest, come yourself”
  • Cic., legg. 3,8: Is iuris ciuilis custos esto, “let him (= the praetor) be the guardian of civil right”.
  • 7.4.6. The tenses of Perfectum

The different tenses of the Perfectum add to the morphological unit of Perfectum the same morphological units as the tenses of the infectum, but these are represented by a minus great number of morphological segments, each of these units being represented by only one morph. The perfect, which is actually a Perfectum present, adds to the Perfectum unit only one of the person units, exactly as the infectum present adds to the verb stem only the person units. But it has at least two complications. Firstly, the Perfectum morpheme has sometimes its full form, that is both segments of the morphological combination: /uis/, /sis/, /: … is/, and /R… is/, and sometimes its short form, that is only the first segment of the combination: /u/, /s/, /: …/, /R …/ before vowel and its variant before consonant: /ui/, /si/, /: … i/, /R …i/. This variation of the Perfectum morpheme only occurs in the perfect, when the morpheme is combined only with a person morphological unit, everywhere else it has the full form. Secondly, two of these person morphological units are represented by a specific segment in the perfect, namely 1 Person by the segment , and 2 Person by -tī:

  • amā-u-ī, amā-uis-tī, monē-u-ī, monē-uis-tī, lēg-ī, lēg-is-tī, etc.

the other person units having the same morphs as in the present: -t for the third Person, -mus for the fourth Person, -tis for the fifth Person, and the allomorph -unt for the 6th Person after consonant. These combined data give the following table of terminations:

where -uer-unt is phonetic realization of /uis-unt/. One can ask oneself why it is not /ui-nt/? Because historically when the thematisation rule concerned the -nt 6th Person, it always led to -unt, and to i-t, when the -t 3th Person, like legi-t and leg-unt; therefore */ui-nt/ was not possible, because it would suppose a thematisation in i before -nt.

But the 6th Person has another specific allomorph -ēre, which is combined with the short form of the Perfectum, since it begins with a vowel. It has no connection with [er-unt] the phonetic realization of /is-unt/, because of its long vowel. It is used rather rarely. The classical prose seems to avoid using it. Cicero says that if “he doesn’t criticize scripsere alii rem (“others wrote the thing”), he feels that scripserunt is more regular”. And the editor notices in a footnote: therefore we find in Caesar only forms in erunt. But Cicero adds that “he obeys the use which seeks what is attractive to the ears”4) . This comment seems to mean that the forms in -ēre was somewhat elegant and original. It is perhaps the reason why many perfects in -ēre are found in Sallust, Livy, and Tacitus.

Finally, there is another 6th Person form in -ērunt, which is, as Ernout says,

  • A. Ernout, Morph. hist.,1953, p. 216: “a compromise between -ěrunt and -ēre, maybe artificial, and created by dactylic poets in order avoid a Cretic foot as amāuěrūnt”.

The pluperfect and the future perfect, which are morphologically the Perfectum imperfect and future, insert a tense morphological segment between the Perfectum morpheme and the Person morpheme. These inserted segments are the same as for the esse verb, i. e. -ā- for the Imperfect and / ~ i/ for the Future, or, if we accept James Foley assumption, -bā- for the Imperfect, and /b ~ bi/ for the Future: then /is-ba:/ and /is-b/ are phonetically realized [is-a:] and [is-] like /es-ba:-s/ and /es-bi-s/ are realized [es-a:-s] and [es-i-s] and, after the rhotacismus:

  • - uer-ā-s, -uer-i-s or -ser-ā-s, -ser-i-s, R… -er-ā-s, R …-er-i-s etc., like er-ā-s “you was” and er-i-s “you will be”.

But in the 6th Person, the Future morphological unit replaces the segment by an /i:/ subjunctive segment in order to avoid a confusion with (u)er-unt of the perfect:

  • amā-uer-i-nt, monu-uer-i-nt, lēg-er-i-nt, fe-fell-er-i-nt, etc.

This exchange between the Future and the Subjunctive is comparable with that of the Future 1th Person of the 2th infectum conjugation:

  • leg-a-m “I shall read”, capi-a-m “I shall take”, audi-a-m “I shall hear”

which is like the Subjunctive 1th Person:

  • leg-a-m “I may read”, capi-a-m “I may take”, audi-a-m “I may hear”.

The perfect subjunctive, which is really the Perfectum subjunctive, inserts a -ī- segment between the Perfectum morpheme and the Person morpheme: amā-uer-i-m, amā-uer-ī-s, amā-uer-i-t, amā-uer-ī-mus, amā-uer-ī-tis, amā-uer-i-nt which is the same as that of the verb es-se:

  • s-i-m, s-ī-t, s-i-t, s-ī-mus, etc.

And the pluperfect subjunctive inserts the /se:/ morphological combination which shows the same [se:] and [se] phonetic realizations as in the es-se verb, since it is after a sibilant consonant, as in the verb es-se:

  • amā-uis-se-m, amā-uis-sē-s, amā-uis-se-t, amā-uis-sē-mus, etc.
  • lēg-is-se-m, lēg-is-sē-s, lēg-is-se-t, etc. fe-fell-is-se-m, fe-fell-is-sē-s, etc.
  • 7.5. The passive

While the active traditional category corresponds to neither a morphological unit, nor a morpheme, the passive category always corresponds to a morphological segment; but this passive segment together corresponds to another morphological unit. It is therefore a morphological combination. Thus amā-t-us su-m “I was loved” is opposed to am-or “I am loved” like amā-u-ī “I loved” is opposed to am-ō “I love”; hence there is in -t-us su- the Perfectum. But amā-tus su-m “I was loved” is opposed to amā-u-ī “I loved”; hence there is in -t-us su- the Passive. But there is already in -t-us su- the Perfectum. The only one conclusion is that -t-us su- is a morphological combination which together corresponds to the Passive unity and Perfectum unity. This combination is formed by three segments: firstly, the -t-(us) suffix called passive participle, which comes after the verbal stem, and is followed by a nominative casual ending, singular or plural according the concerned Person; then, the stem of the verb es-se, which forms a syntagmatically independent word, and usually comes after the word which these two segments belong to, but may very well come before this word:

  • amā-t-us su-m “I was loved”, amā-t-ī su-mus “we were loved”, es-t amā-t-us “he was loved”, s-unt amā-t-ī “they were loved”, etc.

The second word may be replaced by , particularly in the 3th person, amā-t-us having then the same meaning as amā-t-us est. It may be sometimes replaced also by the fu-ī perfect: amā-t-us fu-ī instead amā-t-us su-m; so also amā-t-us fu-er-a-m instead amāt-us er-a-m, etc. The first segment of the passive Perfectum morphological combination is a -s- when the verbal stem is ended by a dental like the verb scind-ō, scinde-re “split” or mitt-ō, mitte-re “send”: scis-s-us es-t “he was split”, phonetic realization of /skid-s-us es-t/, with neutralization of the quantity opposition before a voiceless consonant, and the sibilant variant of a dental before a sibilant consonantal; mis-s-us est “he was sent”, phonetic realization of /mit-s-us/, with the sibilant variant of a dental before a sibilant consonantal. The first segment, -t-us as well -sus, quite often lead to a variant of the verbal stem which it is difficult to expect:

  • domi-t-us for dom-ō, i-s, domŭ-ī “subjugate”; cubi-t-us for cub-ō, ā-s, cubŭ-ī “lie down”; moni-t-us for mone-ō, ē-s “warn”; scis-s-us for scind-ō, i-s, sci-cid-ī or scid-ī “split”; mulc-tus or mul-s-us for mulge-ō, ē-s, mulx-ī or mul-s-ī “milk”; etc.

And when the verbal lexeme shows no morphological variant, it shows at least the phonological variants which are caused by the /t/ and /s/ phonemes, as

  • emp-t-us beside em-ō, i-s, ēm-ī “buy”, scrip-t-us beside scrīb-ō, i-s, scrip-sī “write”, ac-t-us beside ag-ō, i-s, ēg-ī “drive”; fix-us beside fīg-ō, i-s, x-ī “fix in”; dīuī-s-us beside dīuid-ō, i-s, uī-s-ī “separate in two”; etc.

In the absence of the Perfectum morpheme, grammars traditionally speak about the passive infectum. Since the term of infectum doesn’t refer to any morphological characteristic, and since only the person endings are particular to the passive forms, these are not morphological combinations which together indicate the person and the passive. If many of them seem to be constituted with a person segment and a segment in r, as o + r for the first person, t + ur for the third person, and nt + ur for the 6th person, and ri + s for the second person, others show only the segment in r, as the first Person unmarked form -r or the -re variant of the second person, or replace the s of the 4th person -mus with an -r, and finally the -minī 5th person form is without segment r and cannot be analyzed. It therefore is better to consider these endings as any morphological combinations, which in the infectum represent the Passive + the Person, while in the Perfectum the morphological combination represented the Passive + the Perfectum. So, in the first person, the Passive + first Person combination is -or only in all indicative presents and futures in -b-:

  • am-or “I am loved”, mone-or “I am warned”, leg-or “I am read”, capi-or “I am taked”, audi-or “I am heard”
  • amā-b-or “I shall be loved’, monē-b-or “I shall be warned”.

And everywhere else, it is only -r:

  • amā-ba-r “I was loved”, monē-ba-r “I was warned”, leg-ē-ba-r “I was read”, etc. leg-a-r “I shall be read”, capi- a-r “I shall be taken”, etc. am-e-r “I would be loved”, amā-re-m, etc.

In the second person, it is either -ris or -re, but only -re in the imperative:

  • amā-ris (amā-re) “you are loved”, monē-ris (monē-re) “you are loved”, lege-ris (lege-re) “you are read”, etc. amā-re “be thou moved”, monē-re “be thou warned”, lege-re “be thou read”, audī-re “be thou heard”, etc.
  • amā-bā-ris ( amā-bā-re) “you were loved”, monē-be-ris (monē-be-re) “you will be warned”, leg-ēbā-ris (leg-ēbā-re), capi-ēbā-ris (capi-ēbā-re), etc.

according to Ernout,

  • A. Ernout, Morph. hist., 1953, p. 122: “Plautus has only nine certain examples of -ris. Terentius only ever uses re. And Cicero systematically distributes the two forms: in indicative present, he uses generally -ris in order to avoid confusion with imperative (and with active present infinitive); but in the indicative imperfect and future, and present and imperfect subjunctives, where the confusion is not possible any more, he seems to prefer -re.”

In third person, it is always -tur, in fourth person, always -mur, in fifth person, always -minī:

  • amā-tur “he is loved”, amā-mur “we are loved”, amā-minī “you are loved”,
  • monē-tur “he is warned”, monē-mur “we are warned”, etc.

But in sixth person, it is -ntur, which becomes untur after a consonant or an i:

  • amā-ntur “they are loved”, amā-ba-ntur “they were loved”; amā-b-untur ” they will be loved”, leg-untur “they are read”, capi-untur “they are taken”, etc.

Apart from these two morphological particularities which combine Passive with Person in the infectum, or with Perfectum in the Perfectum, the Passive uses the same tense and mood units as the active. There are therefore two main infectum conjugations: the 1st conjugation the verbal stem of which is ended by a [- closed] vowel, which includes verbs in ē-re and verbs in ā-re, and the 2nd conjugation whose the verbal stem is ended by a [+ closed] vowel, which includes the verbs in e-re and verbs in -ī-re. And in each of the two conjugations, there is a sub-conjugation which loses its final vowel; for the 1st conjugation, it is the verbs like am-ō, ā-s; for the 2nd conjugation, it is the verbs like dic-ō, i-s.

In the passive perfectum, there are, in front of the five active perfectum conjugations, only two kinds of conjugations: the conjugation which uses the morphological combination /t-(us) es/, and the conjugation which uses /s-(us) es/. The second segment of these morphological combinations is then the same as the verb es-se and is therefore combined with the same tense and mood morphological segments. Grammars, rightly distinguishing the second word of the passive perfectum from the es-se verb, call it an es-se auxiliary, which is a homonym of this verb.

  • 7.6. Deponent verbs

Some verbs always have a passive form, and never an active one. Grammars call them deponent verbs, and define them as verbs

  • Allen & Greenough, 1888, p. 103, “having the forms of the passive voice with an active or reflexive signification”.

From the morphological point of view, they simply are passive verbs, whatever their meaning. It is therefore not a surprise if like the actual passives they have the same infectum conjugations:

  • uere-or “fear” like mone-or, and imit-or, imitā-rī “imitate” like am-or, amā-rī, for the 1st infectum conjugation
  • experior, experī-rī “attempt” like audi-or, audī-rī; pati-or, pate-ris “undergo” like capi-or, cape-ris, and ūt-or, ūte-ris “use” like leg-or, lege-ris, for the 2nd infectum conjugation.

and the same -t(us) es- and -s(us) es- perfectum conjugations:

  • imitā-t-us su-m, ueri-t-us su-m, exper-t-us su-m
  • pas-s-us su-m “I underwent”, ū-s-us su-m “I used”.

Some verbs have this passive forms only in the perfectum, and not in the infectum. For example, the verbs aude-ō “dare”, gaude-ō “rejoice”, sole-ō “be wont”, and fīd-ō, fide-re “trust”, which are conjugated, the first three, like verbs of the 1st infectum conjugation, and, the last, like a verb of the 2a infectum conjugation, have as perfect respectively au-s-us sum “I dared”, gāuī-s-us su-m “I rejoiced”, fī-s-us su-m “I trusted”, and soli-t-us su-m “I was wont”. They are therefore called semi-deponent verbs.

  • 7.7. Noun and adjective forms

Some verbal forms are without a person morphological segment and called non-personal forms. These verbal forms often have a casual or originally casual segment instead the personal segment. Grammars speak then about noun and adjective forms of a verb. That does not mean the verb has become a noun or an adjective; it still is a verb, but it fulfills some functions which ordinarily are a matter for a noun or adjective.

  • 7.7.1. The infinitive, which properly is an abstract noun denoting the action or state expressed by the verb, differs however from other abstract nouns, because 1) it seems to admit some tenses, 2) it is modified by adverbs, and not by adjectives as nouns are, 3) it is constructed with verb complements, and not with noun complements. But, even though it cannot be declined, it functions as a subject and a verb complement, i. e. as a nominative and an accusative.

It corresponds to a morphological segment /se/, which could be, if its final e is original, an old instrumental, or if its final e goes back to an i, an old locative. In the infectum, it is phonetically realized [re], because of the rhotacismus:

  • monē-re and amā-re for the 1st conjugation
  • audī-re, cape-re and dīce-re for the 2nd conjugation;

but in the perfectum, the signifier of which always ends with a sibilant consonant, is realized [se]:

  • amā-uis-se, monŭis-se, lēg-is-se, fe-fell-is-se

and in the infectum of the es-se verb and its compounds:

  • ad-es-se, ab-es-se, po-s-se phonetic realization of /pot-s-se/.

Since the Infinitive morphological unit replaces the Person morphological unit, it is not surprising to see that the Infinitive is combined with the passive in the infectum, the morphological combination of Infinitive and Passive -rī alternating with in the verbs in ĭ of the 2nd infectum conjugation:

  • amā-rī “be loved”, monē-rī “be warned”, audī-rī “be heard”
  • leg-ī from leg-ō, i-s, e-re; cap-ī from capi-ō, i-s, e-re (we shall notice the likeness between cap-ī = /kapi-i:/ and leg-ī = /leg-i:/).

In the passive perfectum, the -se infinitive segment is added to the passive and perfectum morphological combination -t(us) es- which agrees in gender and number with the accusative subject:

  • amā-t-um esse “(him) to have been loved”, amā-tam es-se “(her) to have been loved”, amā-t-ōs es-se “(them) to have been loved”
  • moni-t-um esse “(him) to have been warned”, lec-t-um es-se “(him) to have been read”, audī-t-ōs es-se “(them) to have been heard”
  • fal-s-um esse “(him) to have been deceived”.
  • 7.8. Gerund and supine

The gerund is a verbal noun, which like the infinitive expresses the action or state signified by the verb in the form of a verbal noun. It corresponds in meaning to the English verbal noun in -ing. It adds to the stem the morphological segment -nd- in the 1st infectum conjugation or -end- in the 2nd conjugation:

  • ama-nd-ī “lov-ing-Gen., mone-nd-ī “warn-ing-Gen.”
  • leg-end-ī “read-ing-Gen.”, capi-end-ī “taking-Gen.”, audi-end-ī “hear-ing-Gen.”

It exists only in oblique cases, and completes, as it were, the declension of the active present infinitive, which replaces it as the nominative: scribe-re es-t util-e “writing (to write) is useful”, in front of ars scribendi “the art of writing”. That corresponds to the following table:

infinitive gerund supine
NOM. amā-re
ACC. amā-re (ad) ama-nd-um amā-t-um
GEN. ama-nd-ī
DAT. ama-nd-ō amā-t-ū (-uī)
ABL. ama-nd-ō amā-t-ū

The supine is also a verbal noun with the segment -t- or -s-, which is declined according to the fourth declension. It is found only in the accusative ending t-um or s-um, and the dative or ablative ending -t-ū or -s-ū:

  • ue-:-n-i-t spectā-t-um “he came to see”; e-ō lū-s-um “I go to play”
  • mirā-bil-e dic-t-ū “wonderful to tell”, mirā-bil-e uī-s-ū “wonderful to see”.

Plautus distinguishes even the Dative in -uī :

  • Plaut.,Bacch. 62: istaec lepida sunt memoratui “these things are agreeable to say”

from the Ablative in :

  • Men. 288: opsonatu redeo “I got back from shopping”.
  • 7.9. Participles

Participles change the action or state expressed by a verb into an element of the NP qualification and make with this meaning an adjective which is declined in all genders and numbers like any adjective.

  • Present participle

For the active or deponent verbs, the so-called present participle, which has the same meaning as the English participle in -ing, adds to verbal stem of the 1st infectum conjugation a morphological segment -nt- or -ent- in the 2nd conjugation: ama-n-s (phonetic realization of /ama:-nt-s/, because of the sibilant alternation of /t/ before a sibilant, and the simplification of the final geminates), nt-is, nt-ium “loving”, from amā-re; imita-n-s, nt-is, nt-ium “imitating” from imitā-rī; mone-n-s, nt-is, nt-ium “warning” from monē-re; uere-n-s, nt-is, nt-ium “fearing” from uerē-rī leg-en-s, nt-is, nt-ium “reading” from lege-re; loqu-en-s, nt-is, nt-ium “speaking” from loqu-ī; capi-en-s, nt-is, nt-ium “taking”, from cape-re; pati-en-s, nt-is, nt-ium “undergoing” from pat-ī; audi-en-s, nt-is, nt-ium “hearing” from audī-re; parti-en-s, nt-is, nt-ium “dividing” from partī-rī.

  • Past passive participle

This participle in -nt- doesn’t exist in the passive voice, which has only a participle in t-us or s-us traditionally called the perfect passive participle or “past passive participle”.

  • amā-t-us “moved”, mis-s-us “sent”, etc.

Grammars “usually teach that in Latin there is, in active voice, a present participle (amans “loving”), but no past participle (cf. fr. ayant aimé “having loved”), in the passive voice, a past participle (amā-t-us “fr. ayant été aimé, Engl. having been loved”), but no present participle (cf. fr. étant aimé “being loved”), but in the deponent voice, a present participle (imitans “imitating”) and a past participle (imitā-t-us “fr. ayant imité , engl. having imitated”). That is true, if we judge it from the most frequent translations into French of the different Latin participles. But that is not true in the Latin system. Because what we traditionally call the Latin present participle and the perfect participle has really no temporal or aspectual value. There is, strictly speaking, neither present participle nor past participle in Latin. There is only a simple participle, which uses the -nt- morphological segment in the so-called active voice, and in the passive, the -t(us) or -s(us) morphological segments. … As to the deponents which use the two segments, these do not really have an opposite contrast meaning, contrary to what should be expected.” (according to Touratier, 1994, p. 161)

  • Future participle?

Grammars, which wrongly believed that a Latin participle could be present and past, identified the verb derived adjectives in -tūr-us or -sūr-us as any future participles. But if there is no present nor past participle, why would there be a future participle? It is rather an adjective which expresses what is likely or intended or about to happen, and differs from the future tense by this meaning. And until the end of the Republican period, this so-called future participle was only used as predicative adjective with the es-se verb.

  • Cic., de orat. 1,223: eorum apud quos aliquid aget aut erit acturus mentis sensusque degustet “he ought to examine the thoughts and feelings of those before whom he will plead or intend to plead any cause”.

It is only from Vergil and Livy that the adjective in -tūr-us, is usual without the verb sum¸ and accorded with a subject or a complement:

  • Verg., Aen. 2,660: Periturae addere Troiae Teque tuosque iuuat , “it delights you to add yourself and your family to Troy destined for the ruin”.
  • Future infinitive?

If the so-called future participle is nothing but a verbal adjective, is it possible to speak, as grammars do, about a future infinitive? Grammars, which think that the perfectum infinitive is a past infinitive, and the infectum infinitive is a present infinitive, do not hesitate to speak about a future infinitive. This future infinitive is said to have been formed, in the active voice, by the future participle in the accusative combined with the infinitive of the verb es-se:

  • amā-tūr-um (-am, -um) es-se “to be about to love”, moni-tūr-um (-am, -um) es-se “to be about to warn”, lec-t-ūr-um (-am, -um) es-se “to be about to read”, etc.

In the passive, it is said to have been formed by the supine accusative combined with the impersonal infinitive of the ī-re verb: amā-t-um ī-rī “to be about to be loved”, moni-t-um ī-rī “to be about to be warned”, lec-t-um ī-rī “to be about to be read”, etc. But, in the deponent verbs, the future infinitive is like in an active verb:

  • Imitā-tūr-um (-am, -um) es-se “to be about to imitate”.

In actual fact, if there is no future participle, there is no future infinitive either. And if there is no future infinitive, we understand that the so-called future infinitive can be constructed, like every predicate adjective, with the perfectum of the esse verb, as in

  • Cic., Mil. 46: Dixit … P. Clodium illo die in Albano mansurum fuisse, “He said that P. Clodius had intended to remain that day at his Alban villa”.

As for the so-called passive future infinitive, it is better to consider it as “a periphrastic locution consisting of a supine and the infinitive of the ī-re verb which is put into the impersonal passive” (Ernout & Thomas, p. 325). And the noun in the accusative that appears in this construction “is not the subject of the infinitive, but rather the complement of the supine” (Ernout & Thomas, p. 325), which explains why the verb is invariable. So,

  • Ter., Hec. 39-40): in rumor uenit datum iri gladiatores ,“a rumor spread that gladiators were about to be exhibited” (Henry Thomas Riley, in: New York: Harper and Brothers, 1874) If the translation is semantically exact, it does not allow understanding the syntactical construction, which must be the same as in
  • Sall., Cat. 30: fuere ciues qui rem publicam perditum irent , « there were citizens who went about to ruin the Republic ».

Here the supine in the accusative is used after a verb of movement to express a purpose and it takes an object in the accusative; hence, the following translation:

  • “a rumor spread that people were about to exhibit gladiators” or perhaps“that some one goes (it is gone) about to exhibit gladiators”.
  • Gerundive

There is another verbal adjective; it is traditionally called the gerundive. It shows a morphological segment -nd- in the 1st infectum conjugation, and -end- in the 2nd conjugation, which is the homonym of the Gerund, since it means firstly passive and secondly obligation, necessity or propriety, when it is used as an adjective with the es-se verb :

  • audiend-us es-t “he must be heard”, capi-end-us est “he is to be taken”, leg-end-us es-t “he must to be read” ama-nd-us e¬s-t “he is to be loved”, monendus es-t “he is to be advised”.

As for the so-called exchanges between the Gerund of cupidus uidendi urbem and the gerundive of cupidus urbis uidendae, they really are two different constructions which denote the same extralinguistic reality, but do not show it in the same way: in the first case, it is a man who loves the action of visiting Rome, in the second, it is a man who loves Rome which he must visit. It is therefore different.

  • 7.10. Irregular verbs

The grammars call regular verbs the verbs of

1) the 1a conjugation that, like am-ō, ā-s, amā-u-ī, amā-t-um, ending in ā-, have /uis ~ u/ as Perfectum segment, and t-um as supine

2) the 1st conjugation that, like dele-ō, ē-s, delē-u-ī, delē-t-um, ending in ē-, have /uis ~ u/ as Perfectum segment, and t-um as supine; but, since delē-re is the only one that keeps its final ē-, the common type of the 1st conjugation is, like mone-ō, ē-s, monŭ-ī, moni-t-um, has a variant in ŭ before the /uis ~ u/ Perfectum segment (the monŭ-ī traditional spelling corresponding to the [monu-w-ī] pronunciation) and in ĭ before the -t-um supine segment.

3) the 2a conjugation that, like teg-ō, i-s, tege-re, tex-ī, tec-t-um “cover”, or dīc-ō, i-s, dīce-re, dix-ī, dic-t-um “say”, has /sis ~ s/ as Perfectum segment and the variant without a final vowel before the -t-um supine segment

4) the 2nd conjugation in ī that, like audi-ō, ī-s, audī-re, audī-u-ī, audī-t-um, has /uis ~ u/ as Perfectum segment, and t-um as the supine.

All the other verbs are more or less irregular. We can distinguish two kinds of irregularity. There is the irregularity of the verbs that show some allomorphs in the context of some tense series, e.g. in the perfectum or in the passive. When the irregularities of their principal parts are known, these verbs are conjugated according to the regular conjugation pattern. There are some much more irregular verbs the conjugation itself of which is irregular.

  • 7.10.1. The verb es-se and its compounds
  • The verb es-se: The first particularity is not an irregularity: the apparent alternation /es- ~ er-/ is actually a phonological variation called the rhotacismus.

If the signifier of this verb is /es-/, it will be realized [es] before a consonant; hence es-t “he is”, es-tis “you are”, and the imperfect subjunctive es-sem, es-sē-s, es-se-t; but before a vowel, it will be realized [er]; hence, in future er-ō, er-i-s, er-i-t, … er-unt, and in imperfect er-a-m, er-ā-s, er-a-t, er-ā-mus, etc. The first and serious irregularity of the verb es-se is the fact that it has no allomorph in i before the consonantal initial of the person segments: es-t “he is” in front of legi-t “he reads”, e-s “you are”, realization of /es-s/ in front of legi-s “you read”, and es-tis “you are” in front of legi-tis “you read”. But before the nasal initial of the same segments, there is an u thematization:

  • su-m “I am”, su-mus “we are”, s-unt “they are”,

and second irregularity, the verb shows an /s/ variant instead of the /es/ morph of the other persons, and so in the subjunctive:

  • s-i-m, s-ī-s, s-i-t, s-ī-mus, etc.

James Foley suggested the following explanation: if we assume that the verb es-se has /s/ as no-marked morph, then /es/ is an allomorph with an e prosthesis, according to the rule:

  • # Consonant + Consonant → # e Consonant + Consonant;


  • /s-t/ → est, /s-tis/ → es-tis

and so, in imperfect subjunctive:

  • /s-se :-s/ → es-sē-s, es-se-m, etc.

As for the imperfect and future, he certainly derived [era:s] from /esa:s/ and [er-o:] from /es-o:/, but he saw in es-bā-s “the antecedent of es-ā-s” (p. 1967, 65)5) , and in es-b-ō the one of es-ō, which had the advantage of simplifying the description of the Imperfect morpheme, and explaining the prosthesis from s-bā-s, which implies a rule of a b suppression in es-bā-s6)) . This rule

  • s + b → s or b → Ø / #s ─

is not without likelihood, since the sb phonetic group does not exist in Latin, which is a Latin particularity, while this group is possible in Greek in the initial as well as intervocalic position (cf. sbšnnumi “embrace”, pršsbuj “old”, prosb£llw “throw towards”). Thus, if we assume the following rules:

1. Introduction of an epenthetic u, when /s/ precedes a person morpheme starting with a nasal consonant, which cannot follow it: thus /s-m/ → su-m or s-um, /s-mus/ → su-mus or s-umus, /s-nt/ → s-unt

2. Prosthesis of an e vowel, when /s/ is followed by another consonant that is possible in Latin ; thus /s-tis/ → es-tis, /s-s/ → es-s

3. Simplification of the s geminate consonant; thus es-s# → es

4. s + bs, because the sb phonetic group is not viable in Latin7). Rhotacismus, for the intervocalic /s/ phoneme just before or after a morpheme boundary, we can account for erās by the following derivation:

  • Foley, p. 65:
    “s + bā + s
    ─ (the 1 rule does not apply)
    es + bā = s (the 2 rule, prosthesis)
    ─ (the 3 rule does not apply)
    es + ā + s (the rule 4 cancels the b)
    er + ā s (the 5 rule, rhotacismus)
    erās” .

Therefore it is possible to say the unmarked signifier of the es-se verb is /s-/, /es-/ simply being an allomorph of this /s-/.

The principal parts of the es-se verb being:

  • su-m, es, fu-ī, fu-tūr-us, es-se

it seems that this verb has also a fu- allomorph that appears in the perfectum with a morphological segment /uis ~ u/ rather than a segment / … is ~ /: fu-ī, fu-istī, fu-i-t, fu-i-mus, etc. because fu-ī was the spelling of [fuwi:], which corresponded to [fu-w-i:], and in the so-called future infinitive:

  • fu-tūr-um, -am, -um es-se “to be about to be”.

It shows another free variant /fo/, which appears in the Imperfect Subjunctive:

  • fo-re-m, beside es-se-m, fo-rē-s beside es-sē-s, etc.

and in the so-called future infinitive fo-re, where there is no morphological segment similar to the one -tūr- of fu-tūr-um es-se.

  • verbs posse and prōdesse seem irregular only provided we do not know the Latin phonological system, because possum is the phonetic realization of /pot-su-m/, and prōsum of /pro:d-su-m/, where the /t/ and /d/ apico-dental phonemes receive a sibilant phonetic realization before a sibilant consonant, and the sibilant geminate is simplified after a long vowel.

In these conditions, the apparent alternations are no longer surprising. They follow the one of the es-se verb. When it begins with a vowel, there is no change

Pot-es, pot-est, pot-estis
prōd-es, prōd-est, prōd-estis
pot-er-a-m, pot-er-ā-s, pot-er-a-t, etc.
pot-er-ō, pot-er-i-s, pot-er-i-t, etc.
prōd-er-a-m, prōd-er-ā-s, prōd-er-at, etc.
prōd-er-ō, prōd-er-i-s, prōd-er-i-t, etc.

When it begins with an s¸ the phonological rule is applied:

  • pos-su-m, pos-su-mus, pos-s-unt
    prō-su-m, prō-su-mus, prō-s-unt.

The only irregularity is, in the verb pos-se, the extension of the /s/ signifier to the imperfect subjunctive and the infinitive, or rather, according to Foley’s assumption, the resort to the unmarked signifier:

  • /pot-s-sē-s/ and /pot-s-se/

with moreover the simplifying of the [sss] group, which does not exist in Latin8) , and the potu- variant of the verbal stem in the perfectum:

  • pot-u-ī “I could” [potu-wi:] , pot-uis-tī, pot-ui-t, etc.

But the prōd-es-se verb is entirely regular. Its principal parts are:

  • prō-su-m, prōd-es, prō-fu-ī, prō-fu-tūrus, prōd-es-se

and we find, in all the tenses, the forms of the verb es-se:

  • prō-su-m, prōd-er-a-m, prōd-er-ō, prō-s-i-t, prōd-es-se-m, etc.

with only prō- before the forms beginning with s:

  • prō-sum, prō-sit, etc., because of the processing of apicodentals before the sibilant [s].
  • The compounds ab-es-se or ad-es-se are not at all irregular. The final consonant of their prefix is only affected by the neutralization of the voicing oppositions before the /s-/ allomorph of the verb es-se. But, by analogy, it is not spelt; hence the form written absum and adsum, despite a pronunciation [apsum] and [assum].

As far the de-es-se compound verb, the final vowel of its prefix shows the neutralization of the quantity opposition before the /es-/ allomorph of the es-se verb; hence the variation between dē-su-m and de-es-t. And the inter-es-se verb shows no phonological particularity.

  • 7.10.2. The verb fer-ō and its compounds

The great irregularity of this verb is the fact that it is a member of the 2a conjugation, but without having the allomorph in i before the morphological segment beginning with a apico-dental consonant, i. e. /s/, /t/ and /r/. It is conjugated like the verb leg-ō, i-s:

  • fer-ēba-m, fer-ēbā-s like leg- ēba-m, leg-ēbā-s, etc. in the imperfect
    fer-a-m, fer-ē-s like meg-a-m, leg-ēs, etc. in the future
    fer-a-m, fer-ā-s like leg-a-m, leg-ā-s, etc. in the subjunctive
    fer-en-s, -ent-is like leg-en-s, ent-is, in the participle


  • fer-s, fer-t, fer-tis unlike legi-s, legi-t, legi-tis in the present
    fer-re phonetic realization of /fer-se/, unlike lege-re, phonetic realization of /legi-se/, in the infinitive
    fer-re-m, fer-rē-s phonetic realizations of /fer-se:-m/, /fer-se:–s/, unlike lege-re-m, lege-rē-s phonetic realizations of /legi-se:-m/, /legi-se;-s/, in the imperfect subjunctive.

The situation is the same of course in the passive:

  • fer-ris, fer-tur in the present, fer-rī in the infinitive, fer-re-r, fer-rē-ris in the imperfect subjunctive.

On the other hand, the verb has the allomorph in i when it is found before a morphological segment that does not begin with an apico-dental consonant:

  • feri-mus “we bear”, feri-mur “we are born”, feri-minī “you are born”.

The second irregularity of this verb follows from its principal parts:

  • fer-ō, fer-s, tul-ī, lā-t-um, fer-re.

It shows therefore an allomorph /tul-/ in context of the Perfectum morpheme not preceded by the Passive, the Perfectum corresponding then to a morphological segment /…is ~ /, and an allomorph /la:/ in the context of the Passive going with Perfectum:

  • tul-ī “I bore”, tul-i-t “he bore”; lā-t-us su-m “I am bore”, lā-t-us er-a-m “I was bore”, lā-tūr-us su-m “I am about to bear”.

The compounds of fer-re add to these particularities the phonological particularities that concern the possible consonantal final of their prefix, but are not always spelled:

  • ad-fer-ō “bring” pronounced [affero:], ad-fer-s, ad-tul-ī pronounced [attuli:], ad-lā-t-um pronounced [alla:tum]
    ef-fer-ō “carry or bring out or away”, ef-fer-s, ex-tul-ī, ē-lā-t-um
    dif-fer-ō “scatter, disperse”, dif-fer-s, distul-ī, dī-lā-t-um
    au-fer-ō “carry or fetch away”, au-fer-s, abs-tul-ī, ab-lā-t-um
    of-fer-ō “bring to a place”, of-fer-s, ob-tul-ī, ob-lā-t-um
    in-fer-ō “carry or convey into a place”, in-fer-s, in-tul-ī, il-lā-t-um
    re-fer-ō “bring back or again”, re-fer-s, ret-tul-ī, re-lā-t-um
    suf-fer-ō “submit to”, suf-fer-s, sus-tul-ī, sub-lā-t-um9) .
  • 7.10.3. The verb uol-ō and its compounds
  • The verb uol-ō whose principal parts are
  • uol-ō, uī-s, uolŭ-ī, uel-le has the particularity of not knowing the allomorph in i uul-t, uul-tis, uel-le phonetic realization of /uel-se /, uel-le-m phonetic realization of /uel-se:-m/ but nevertheless it follows the 2a conjugation, like fer-ō, while it shows some particularities of the su-m conjugation. It is certainly conjugated like fer-ō or leg-ō, with the imperfect uol-ēba-m, uol-ēbā-s and the future uol-a-m, uol-ē-s, like fer-ēba-m and fer-a-m; but in the subjunctive it gives uel-i-m, uel-ī-s etc. like s-i-m, s-ī-s etc. and it shows the same variant en u- in the present fourth person uolu-mus, like su-mus, and unlike feri-mus. To this morphological irregularity is added a phonological particularity that shows that the basic allomorph of the verb is uel-. After an initial u the palatal phoneme /e/ is realized as a velar o before a explosive velar [ł], an u before an implosive [ł], i. e. an [ł] followed by a consonant that is not a second lateral, but an e before a palatal [l]. Thus, the lateral being velar before every vowel except /i/ or every consonant except /l/, /uel/ is realized [woł-] in
  • uol-ō “I want”, uolu-mus “we want”, uol-unt “they want”, uol-ēba-m “I wanted”, uol-ēbā-s, etc. uol-a-m “I shall want”, uol-ē-s, etc. uol-en-s, uol-ent-is “wanting”

[wuł-] in

  • uul-t “he want”, uul-tis “you want”

but [wel-] in

  • uel-i-m, uel-ī-s, etc. and uel-le-m, uel-lē-s, etc.

In the second Person of the present infectum, the morphological segment uel- is replaced by a segment uī-, hence uī-s “you want” (cf. in-uī-tus “not wishing, unwilling”), probably because the consonantal group [ls#] is almost impossible in Latin, puls, lt-is “a kind of porridge” quoted by Touratier (2005, p. 127, n. 2) and uls “on the farther side of” being the only two examples found in the data basis itinera electronic. In perfectum it is the allomorph uolu- that is used with the perfectum morphological segment /uis ~ u/:

  • uolŭ-ī “I wanted” corresponding to [wołu-w-i:], uolŭ-is-tī to [wołu-wis-ti:], etc.
  • The compounds

The two compound verbs /no:n uel/ “be unwilling, will not” and /ma:-uel/ “be more willing, prefer”, the principal parts of which are:

  • nōl-ō, nōn uī-s, nōlŭ-ī, nōl-le
    māl-ō, mā-uī-s, mālŭ-ī, māl-le

exactly follow the same conjugation as the simple verb, but show moreover an allomorph that cancels the begin of the simple verb every time that it would receive an intermediate realization, i. e. [e] or [o], hence māl-ō, realization of /ma:-uol-ō/, māl-unt, realization of /ma:-uol-unt/, māl-ēba-m, realization of /ma:-uol-e:ba:-m/, māl-a-m¸ realization of /ma:-uol-a:-m/,
māl-i-m, realization of /ma:-uel-i-m/, māl-le-m¸ realization of /ma:-uel-le:-m/, mā-le, realization of /ma:-uel-le/.
nōl-ō, realization of /no:n uol-o:/, nōl-unt, realization of /no:n uol-unt/, nōl-ēba-m, realization of /no:n uol-e:ba:-m/, nōl-a-m¸ realization of /no:n uol-a:-m/,
nōl-i-m, realization of /no:n uel-i-m/, nōl-le-m¸ realization of /no:n uel-le:-m/, nō-le, realization of /no:n uel-le/, beside these three forms:

  • mā-uī-s, mā-uul-t, and mā-uul-tis
    nōn uī-s, nōn uul-t, nōn uul-tis

which use another allomorph than uel- or uol-.

  • 7.10.4. The verb e-ō, ī-re

Having the following principal parts:

  • e-ō, ī-s, i-ī (ī-u-ī), ĭ-t-um, ī-re

it is a member of the 2nd conjugation, like its morph ī- shows in the second person of the so-called present (ī-s) and in the infinitive (ī-re), but it uses the imperfect and future morphological segment of the 1st conjugation:

  • ī-ba-m “I went”, ī-bā-s, ī-b-ō “I shall go”, ī-bi-s, etc.

In the participle, it uses the expected segment -ent- of the 2nd conjugation, for the singular nominative and neuter accusative, but an allomorph -unt- for the other cases:

  • i-en-s “going (Nom. masc. or neut.)”, e-unt-is “going (Gen.)”,…, e-unt-em (Acc. masc.), but i-en-s “going (Acc.neut.)”, etc.

In low Latin, the analogical genitive i-ent-is was created (cf. ientibus in C.I.L. VI, 10241, 12).

The gerund has an allomorph -und- instead of the expected form -end-:

  • ad e-und-um “in order to go”, tempus e-und-ī “time to go”;

Beside these conjugation anomalies, the verb ī-re shows an allomorph e- that is phonologically conditioned: it appears before a vowel, different from e. Thus e-ō “I go”, e-unt “they go”, e-unt-is “going (Gen.)”, e-und-um , e-a-m “I may go”, e-ā-s


  • i-en-s “going (Nom.)”, phonetic realization of /i:-ent-s/. Before a consonant, it is therefore the allomorph /i:/:
  • ī-ba-m, ī-b-ō, ī-re, ī-re-m,

which receives a short realization before a vowel, as in the perefect i-ī, the pluperfect i-er-am or the future perfect i-er-ō. The verb ī-re has another allomorph /i/, which is not an allophone of /i:/, in the supine ĭ-t-um, and the so-called future participle ĭ-tūr-us.

In the perfectum, the verb ī-re regularly uses the morphological segment / … is/:

  • i-ī, i-is-tī, i-i-t, i-i-mus, i-is-tis, i-er-unt phonetic realization of /i:-is-unt/

without the contraction of ii which would lead to a confusion with the present i-t, i-mus; ii before s is regularly contracted to ī: as īsse (= /i:-is-se/), īstī (= /i:-is-ti:/), because a confusion with the present is not possible. But a rare perfect in -uis- was created particularly in poetry (Cat., 66,12: ī-uer-a-t), maybe in order to avoid the tribrach ĭ-ěr-ăt.

  • The compound verbs ad-eō “approach” and in-eō “enter” are transitive and can be inflected in the passive:
  • ad-eor, ad-ī-ris, ad-ī-tur, ad-ī-mur, ad-ī-minī, ad-e-untur
    impf. ad-ī-ba-r, fut. ad-ī-bo-r, subj. ad-e-a-r, impf. Subj. ad-ī-re-r
    inf. ad-ī-rī, part. ad-i-t-us, gerundive ad-e-und-us
    perfectum ad-i-t-us su-m, er-a-m, er-ō, s-i-m, es-se-m.

The simple verb e-ō has also some passive forms, but used impersonally: ī-tur “some one goes (it is gone)”, i-t-um est “some one went”, and ī-rī (cf. p. 76 the so-called passive future infinitive). Plautus has a form of passive infinitive irier10)) .

The verb uēne-ō, ī-s, i-ī or ī-u-ī “be sold” (for uēnum ī-re “go to sale”), which is used as a passive of uend-ō, i-s, did-ī, di-t-um, e-re “sell”, follows the conjugation of e-ō, and has also several forms in the passive, as the infinitive ueniri11).

The verbs neque-ō, ī-s, i-ī or ī-u-ī, ī-re “be unable to”, and que-ō, ī-s, i-ī or ī-u-ī, ī-re “be able to”, follow also the conjugation of e-ō, which they are compound (neque-ō is probably derived from an impersonal phrase *neque itur “it does not go well”, and que-ō by backformation from neque-ō).

  • 7.10.5. The verb fie-rī
  • “1) be made → 2) become”

This verb, the principal parts of which are:

  • fī-ō, fī-s,fac-t-us su-m, fierī

is the passive of

  • faciō, i-s, fēc-ī, fac-t-um, face-re “make”

which is regular. But it has the imperative fac, and the perfectum future fax-ō, and perfectum subjunctive fax-i-m, besides the regular forms fēc-er-ō and fēc-er-i-m.

As for fī-ō, its infectum system is regular and belongs to the 2nd conjugation but the subjunctive imperfect is fierem, and the infinitive fierī. This appearent irregularity is very easily explained if we suppose that this verb shows, in infectum, an allomorph /fi:i/ and in perfectum an allomorph fac-. And the irregularity is the fact that the signifier is associated with the signified “be made”, which is really the passive of “make”, and thus it needs no Passive morphem, which nevertheless appears in the infinitive fie-rī. In the perfectum, it is the supine of the verb faci-ō “make” that is used; in order to correspond to the same signified, it needs the Passive morphem and thus the morphological combination with the perfectum or the person unit.

The irregularities mentioned by grammars between the ī de fīō which doesn’t become short before a vowel, and the expected short ǐ of fierem and fierī, which don’t seem to correspond to the 2nd conjugation, fie-rem corresponding to cape-rem, but with one additional i, and fie-rī to audī-rī, but with one additional e. If we admit that this verb has an allomorph /fi:i/ in the infectum, then in /fi:i-o:/ the short i, which is between two vowel has a consonantal phonetic realization [j], and the /i:/ which is not found before a vowel has no reason to become short; hence fī-ō. And so, for fī-unt, fī-ēba-m, fī-a-m, fī-ē-s, etc. But in the imperfect subjunctive /fi:i-se:-m/, the short i is before a consonant and therefore has a vocalic realization; but this consonant /s/ being between two vowels and after a morpheme boundary is phonetically realized [r], which leads the neutralization of the /i/ ~ /e/ opposition, and therefore an intermediary realization for the previous vowel. And the long ī before this vowel receives a bisegmental realization [ij] because the neutralization of the quantity oppositions, hence [fijerem] spelled fierem. And so, /fi:i-ri:/ with the not necessary and anomalous morphological combination -rī of passive infinitive makes [fijeri:], which is fully phonologically normal.

As for the other persons of the so-called present, fī-s corresponds to /fi:i-s/ which is realized [fi:-s] spelled fis, because the obligatory contraction into a long ī of two i not separated by a morpheme boundary; fit corresponds to /fi:i-t/; but beside the contraction into a long ī, there is neutralization if the quantity oppositions before word final consonant other than s, hence the phonetic realization [fit].

Most compounds of faci-ō show the morphological alternations /a/ ~ /i/ in an open syllable, and /a/ ~ /e/ in a closed syllable, which are synchronical consequences of the historical phonetic change called apophony, are inflectd in the following way:

  • con-fici-ō, -fici-s, -fēc-ī, -fec-t-um, -fice-re « finish »

and in the passive:

  • con-fici-or, -fice-ris, -fec-t-us su-m, fic-ī.

But a few isolated forms of fī-ō can occur in these verbs: confit “it happens”, dēfit “it lacks”, inter-fī-a-t “let him perish”, inter-fie-rī “to perish”.

Some compounds retain the a¸ and normally have -fi-ō in the passive: as, pate-faci-ō “to make visible”, cale-faci-ō “make hot”, bene-faci-ō “do a service to”, satis-faci-ō “give satisfaction”.

But occasionly occur calfacientur (Vitr. 5,10,1), satisfacitur (Var., Men.,82)

  • 7.10.6. The verb ed-ō, i-s, ēd-ī, ē-s-um, e-re, “eat”

which must not be confused with the verb

  • ēd-ō, i-s, did-ī, di-t-um, e-re “eject, emit”,

has become a regular verb of the 2a conjugation; but besides the forms in i it shows some forms without i but with an initial long ē, whereas the forms of the 2a conjugation start with a short e:

  • ē-s beside edi-s, ēs-se-m beside ede-re-m, phonetic realization of /edi-se:-m/, ēs-se beside ede-re phonetic realization of /edi-se/

which can be descripted by a variant /e:d/ immediately before a phoneme /s/. That would correspond to the diachronic change of *ed-s > *ēs-s (with regressive assimilation and compensatory lengthening of the previous vowel12) ), and finally *ēs-s > ēs 13)), and *ed-tos > *ētstos, ēs-sus14) , ēsus15). But there are also

ēst, ēstis, (the passive ēs-tur) beside edi-t, edi-tis (and edi-tur),

which cannot correspond to /e:d-t/ and /e:d-tis/, and even result from *ed-tis, as they should have become *ēs and *ēsis, and not ēs-t and ēs-tis; but “the analogical process impeded the phonetic process”, as Ernout writes16) . Therefore, it is necessary to assume the existence of an allomorph /e:s-/ phonologically conditioned by a morphem beginning with an apico-dental /t/ or /s/. Then, like the verbs without an allomorph in i (cf. s-i-m and uel-i-m), it uses the subjunctive allomorph in -i-m, beside the expected subjunctive in -a-:

  • ed-i-m, ed-ī-s, ed-i-t, etc. beside ed- a-m, ed-ā-s, ed-a-t, etc.

The two forms concurrently existed until the Augustan age; “Horatius always uses edim, Ovidius edam” (according to Ernout, 19533, p. 184).

So the verb //ed-ō// is conjugated in following way:

PRESENT: ed- ō, edi-s (ēs), edi-t (ēs-t), edi-mus, edi-tis (ēs-tis), ed-unt
IMPERFECT: ed-ēba-m, ed-ēbā-s, ed-ēba-t, etc.
SUBJUNCTIVE: ed-a-m, (ed-i-m), ed-ā-s (ed-ī-s), ed-a-t (ed-i-t), etc.
IMPERFECT SUBJUNCTIVE: ēs-se-m, ēs-sē-s, ēs-se-t, etc. (ed-ere-m, ede-rē-s, etc. being secondary and analogical17) )
IMPERATIVE: ēs, ēs-te
FUTURE IMPERATIVE: ēs-tō, ēs-tō-te, ed-untō
INFINITIVE: ēs-se (ede-re being secondary and analogical)
GERUND: ed-end-ī, d-ō, d-um
SUPINE: ēs-um, ēs-ū

  • 7.10.7. The verb d-ō “give” is regularly conjugated like am-ō, amā-s, but it has an ă everywhere am-ō has an ā, except in the present dā-s and the imperative . Thus it is conjugated in the following way:

in the present: d-ō, dā-s, dăt, dă-mus, dă-tis, dă-nt
in the other tenses and moods: dă-ba-m, dă-b-ō, dă-re, dă-te, dă-tō, dă-re-m, etc.
in the passive: da-ri, da-tur, da-ba-r, da-b-or, da-t-us su-m, da-t-us er-a-m, da-rī, da-t-um es-se, etc.

If the subjunctive d-e-m, d-ē-s, d-e-t is classical, archaic Latin retains subjunctive forms which show the original autonomy of the subjunctive:

  • Plaut., Amph. 72: Duint , Plaut., Capt. 947: ne duīs , Plaut., Aul. 238: ne duās.
  • 7.11. Classified lists of verbs
  • 7.11.1. The first conjugation
  • The 1st conjugation

There are “about 570 verbs in -ē-, among which 180 are simple verbs” (according to Ernout, 1953, p.143)

  • Most of these verbs are inflected like mone-ō “warn”, ŭ-ī, i-t-um:
  • cale-ō “be warm”, debe-ō, habe-o, praebe-ō, tace-ō, etc.
  • some have certainly a perfetum in ŭ-ī, but no supine in i-tum
  • time-ō “fear”, timŭ-ī, ─
    care-ō “lack”, carŭ-ī, ─
    ege-ō “need”, egŭ-ī, ─
    cense-ō “value”, cēnsŭ-ī, cēn-s-um
    doce-ō “teach”, docŭ-ī, doc-t-um
    misce-ō “mix”, miscŭ-ī, mix-t-um
    tene-ō “hold“, tenŭ-ī, ten-t-um
    torre-ō “roast“, torrŭ-ī, tos-t-um
  • Some of these have the same morph in all the form systems:
  • dele-ō “destroy”, dēlē-re, dēlē-u-ī, dēlē-t-um
    fle-ō “weep”, flē-re¸ flē-u-ī, flē-t-um
    ne-ō “sew”, nē-re, nē-u-ī, nē-t-um
    com-ple-ō “fill up”, -plē-re, -plē-u-ī, -plē-t-um
    ex-ple-ō “fill up”, -plē-re, -plē-u-ī, -plē-t-um
    im-ple-ō “fill”, -plē-re, -plē-u-ī, -plē-t-um
    re-ple-ō “fill again”, -plē-re, -plē-u-ī, -plē-t-um.
  • some of these have a reduplication perfectum:
  • morde-ō “bite”, mo-mord-ī, mor-s-um
    pende-ō “hang”, pe-pend-ī, ─
    sponde-ō “pledge”, spo-pond-ī, spōn-s-um
    tonde-ō “shear”, to-tond-ī (tond-ī), tōn-s-um.
  • a number of these has a sigmatic perfectum:
  • ārde-ō, ār-s-ī, ār-sūr-us
    rīde-ō, rī-s-ī, rī-s-um
    suāde-ō “urge”, suā-s-ī, suā-s-um
    haere-ō “cling”, hae-s-ī, hae-s-um
    mane-ō “wait”, man-s-ī, man-s-um
    fulge-ō “shine”, ful-s-ī, ─
    torque-ō “twist”, tor-s-ī, tor-t-um
    indulge-ō “indulge”, indul-s-ī, indul-t-um
    iube-ō “order”, ius-s-ī, ius-s-um
    auge-ō “increase”, aux-ī, auc-t-um
    lūce-ō “shine”, lūx-ī, ─.
  • some have a lengthening perfectum:
  • caue-ō “care”, cāu-ī, cau-t-um
    faue-ō “favor”, fāu-ī, fau-t-um
    foue-ō, “cherish”, fōu-ī, fō-t-um
    moue-ō “move”, mōu-ī, mō-t-um
    uoue-ō “vow“, uōu-ī, uō-t-um
    sede-ō “sit”, sēd-ī, ses-s-um
    uide-ō “see“, uīd-ī, uī-s-um.
  • three are semi-deponent:
  • aude-ō “dare”, au-s-us su-m
    gaude-ō “rejoice”, gāuī-s-us su-m
    sole-ō “be wont“, soli-t-us su-m
  • The 1a. conjugation. There are “about 3620 verbs in -ā-, among which 1800 are simple verbs” (according to Ernout, 1953, p.138)
  • most of these verbs have only two allomorphs, like amā- ~ am-
  • two verbs have a reduplication perfectum:
  • dă-re “give”, de-d-ī, dă-t-um
    stā-re “stand”, ste-t-ī, stā-tūr-us
  • some verbs have a perfectum in ŭ-ī:
  • crepā-re “resound”, crepŭ-ī, crepi-t-um
    cubā-re “be in bed”, cubŭ-ī, cubi-t-um
    domā-re “subdue”, domŭ-ī, domi-t-um
    secā-re “cut”, secŭ-ī, sec-t-um
    sonā-re “sound”, sonŭ-ī, soni-t-um
    tonā-re “thunder“, tonŭ-ī, ─
    uetā-re “forbid”, uetŭ-ī, ueti-t-um.
  • one verb with a lengthening perfectum
  • iuuā-re “help”, iūu-ī, iū-t-um
  • 7.11.2. The 2nd conjugation
  • The 2nd conjugation

There are ─ beside 5 deponents and some derivatives in -ŭri-ō, among which only two are in common usage: par-turī-re “be in labor” from pari-ō “bring forth”, and ēsurī-re “be hungry” from ed-ō “eat” ─ 65 denominatives ī-re (according to Mignot, 1969, Les verbes dénominatifs latins, p. 51)

  • Verbs with a -uis- perfectum:
  • audī-re,
    sepelī-re, sepelī-u-ī, pul-tum “bury”;
    cupi-ō, cupī-u-ī, cupī-t-um, e-re “wish for, desire”;
    sapi-ō, sapī-u-ī, e-re “have a taste, show good sense”,”
  • Verbs in ŭī:
  • aperī-re, aperŭ-ī, aper-tum “open”,
    operī-re, uī, tum “cover”,
    salī-re, uī, tum “jump, leap”;
    rapi-ō, e-re, uī, t-um “take away”.
  • Verbs with lengthening perfectum:
  • uenī-re, uēn-ī, uen-tum “come”;
    fodi-ō, fōd-ī, fos-sum, fode-re “dig”;
    faci-ō, fēc-ī, fac-tum, face-re “make”;
    iaci-ō, iēc-ī, iac-tum, iace-re “throw”;
    fugi-ō, fūg-ī, fugi-tūr-us, fuge-re “flee”.
  • Verbs with sigmatic perfectum:
  • farcī-re, far-s-ī, far-tum (farc-tum) “stuff”;
    haurī-re “drain”, hau-s-ī, haus-tum, this verb shows the difference between the historic rhotacismus, which happened (haurī- < *hausi-), and the synchronic rhotacismus, which doesn’t happen (/hauri:/ alternating with /haus/ in the perfect and supine).
    sarcī-re, sars-ī, sar-tum “patch”,
    sentī-re, sen-s-ī, sen-sum “feel”,
    uincī-re, uinx-ī, uinc-tum “bind”
    quati-o, ─, quas-sum, quate-re “shake”;
    concuti-ō, -cus-sī, -cus-sum, -cute-re “shake”;
    percuti-ō, -cus-sī, -cus-sum,-cute-re “strike forcibly hit”.
  • Verbs with Reduplication perfectum:
  • pari-o, pe-per-ī, par-tum, pare-re “give birth to, bear”.
  • The 2a conjugation

“about 570 simple verbs and 1830 compound verbs, in all: about 2400 verbs” (Ernout, 1953, p. 124)

  • with sigmatic perfectum: stem ending with a bilabial:
  • carp-ō, carp-s-ī, carp-tum “pluck”;
    rēp-ō “creep”, rēp-s-ī, ─ ;
    scalp-ō, scalp-s-ī, scalp-tum “scrape”;
    sculp-ō “carve”, sculp-s-ī, sculp-tum;
    serp-ō “crawl”, serp-s-ī, ─;

stem ending with a voiced bilabial:

  • nūb-ō “marry”, nūp-s-ī, nūp-tum;
    scrīb-ō “write”, scrīp-s-ī, scrīp-tum.

stem ending with a voiced apico-dental:

  • cēd-ō “yield”, ces-s-ī, ces-sum;
    claud-ō “shut”, clau-s-ī, claus-sum;
    dīuid-ō “divide”, dīuī-s-ī, dīuī-sum;
    frend-ō “gnash”, ─, frēs-sum (fres-sum);
    laed-ō “hurt”, lae-s-ī, lae-sum;
    lūd-ō “play”, lūd-ī, lū-sum;
    plaud-ō “applaud”, plau-s-ī, plau-sum;
    rād-ō “scrape”, rā-s-ī, rā-sum;
    rōd-ō “gnaw”, rō-s-ī, rō-sum;
    trūd-ō “thrust”, trū-s-ī, trū-sum;
    uād-ō “go”, uā-s-ī, uā-sum.
    mitt-ō “send”, mī-s-ī (= /mi:t-s-i:/), mis-sum (= /mit-sum/);
    omitt-ō “release from”, omī-s-ī, omis-sum.

stem ending with a velar:

  • dīc-ō “say”, dīx-ī, dic-tum;
    dūc-ō “guide”, dūx-ī, duc-tum;

stem ending with a voiced velar:

  • ang-ō “choke”, ānx-ī, ─;
    cing-ō “bind”, cinx-ī, cinc-tum;
    ēmung-ō “clean out”, -mūnx-ī, -mūnc-tum;
    fīg-ō “fix”, fīx-ī, fīx-um;
    fing-ō “fashion”, finx-ī, fic-tum;
    -flīg-ō “smite”, -flīx-ī, -flīc-tum;
    frig-ō “fry”, frīx-ī, frīc-tum;
    iung-ō “join”, iūnx-ī, iūnc-tum;
    perg-ō “go on”, perrēx-ī, perrēc-tum;
    ping-ō “paint”, pīnx-ī, pic-tum;
    plang-ō “beat”, plānx-ī, plānc-tum;
    reg-ō “rule”, rēx-ī, rēc-tum;
    string-ō “bind”, strīnx-ī, stric-tum ; \\ sūg-ō “suck”, sūx-ī, sūc-tum;
    surg-ō “rise”, surrēx-ī, surrēc-tum;
    teg-ō “shelter”, tēx-ī, tēc-tum;
    ting-ō “stain”, tīnx-ī, tīnc-tum;
    merg-ō “plunge”, mer-s-ī, mer-sum;
    sparg-ō “scatter”, spar-s-ī, spar-sum;
    terg-ō “wipe”, ter-s-ī, ter-sum;

stem ending with a labiovelar:

  • coqu-ō “cook”, cox-ī, coc-tum;
    -stingu-ō “quench”, -stīnx-ī, -stīnc-tum;
    ungu-ō (ung-ō) “anoint”, ūnx-ī, ūnc-tum;

stem ending with ct:

  • flect-ō “bend”, flex-ī, flex-um;
    nect-ō “weave”, nex-ī (nexu-i), nex-um;
    pect-ō “comb”, pex-ī, pex-um;
    plect-ō “braid”, plex-ī, plec-tum .

stem ending with a nasal:

  • cōm-ō “comb, deck”, cōmp-s-ī, cōmp-tum;
    dēm-ō “take away”, dēmp-s-ī, dēmp-tum;
    prem-ō, pres-s-ī, pres-sum “press”;
    prōm-ō “bring out”, prōmp-s-ī, prōmp-tum;
    sūm-ō “take”, sūmp-sī, sūmp-tum.
    temn-ō “despise”, temp-s-ī, temp-tum.
    flu-ō “flow”, flūx-ī, flux-um;
    stru-ō “build”, strūx-ī, strūc-tum.
    trah-ō “drag”, trāx-ī, trāc-tum;
    ueh-ō “draw”, uēx-ī, uec-tum.
    uīu-ō “live”, uīx-ī, uīc-tum.
    ger-ō “carry”, ges-s-ī, ges-tum (there are two allomorphs /geri/ ~ /ges/);
    ūr-ō “burn”, us-s-ī, us-tum.
  • with -uis- or only -is- Perfectum?
  • acu-ō, acu-ī, acū-tum “sharpen”; the present acu-ō corresponds to [akuwo:], which is the phonetic realization of /aku:-o:/ (cf. acū-tum). As for the perfect acu-ī, is it a perfect in /is/, as /aku:-i:/ or /aku-i:/, or a perfect in /uis/? If it was in /uis/, it would correspond to a phonological sequence /aku:-uis-/ or /aku-u-i:/.
  • argu-ō “accuse” (= /argu:-ō/), argu-ī, argū-tum;
    imbu-ō “give a taste of”, imbu-ī, imbū-tum;
    lu-ō “wash”, lu-ī, -lū-tum;
    metu-ō “fear”, metu-ī, metū-tum;
    minu-ō “lessen”, minu-ī, minū-tum;
    statu-ō “establish”, statu-ī, statū-tum;
    su-ō “sew”, su-ī, sū-tum;
    (ex)u-ō “put off”, u-ī, ū-tum;
    tribu-ō “assign”, tribu-ī, tribū-tum;
    (con)gru-ō “agree”, gru-ī, ─ ;
    -nu-ō “nod”, -nu-ī, ─ ;
    spu-ō “spit”, spu-ī, ─ ;
    sternu-ō “sneeze”, sternu-ī, ─ ;
    ru-ō “fall”, ru-ī, rŭ-tum (rui-tūr-us), /ruo:/ would be better than /ruuo:/, because it would explain why the supine is rutum and not *rūtum.

Perfectum in [i:-wis]:

  • arcess-ō “summon”, arcessī-u-ī, arcessī-tum;
    capess-ō “undertake”, capessī-u-ī, ─ ;
    incess-ō “attack”, incessī-u-ī, ─ ;
    lacess-ō “provoke”, lacessī-u-ī, lacessī-tum;
    pet-ō “seek to obtain”, petī-u-ī or peti-ī, petī-tum;
    quaer-ō “seek”, quaesī-u-ī or quaesi-ī, quaesī-tum;
    rud-ō “bray”, rudī-u-ī, ─ ;
    scisc-ō “decree”, scī-u-ī, scī-tum;
    ter-ō “rub”, trī-u-ī, trī-tum.
    sin-ō “permit”, sī-u-ī, si-tum.

perfectum in [u-wis]:

  • al-ō “nourish”, alu-ī, al-tum (ali-tum);
    col-ō “dwell, till”, colu-ī, cultum;
    compēsc-ō “restrain”, compēscu-ī, ─ ;
    cōnsul-ō “consult”, cōnsulu-ī, cōnsul-tum;
    -cumb-ō “lie down”, -cubu-ī, cubi-tum;
    deps-ō “knead”, dessu-ī, deps-tum;
    frem-ō “roar”, fremu-ī, ─ ;
    gem-ō “groan”, gemu-ī, ─ ;
    gign-ō “beget”, genu-ī, geni-tum;
    met-ō “reap”, messu-ī, -messum;
    mol-ō “grind”, molu-ī, moli-tum;
    occul-ō “hide”, occulu-ī, occul-tum;
    pōn-ō “put”, posu-ī, posi-tum;
    ser-ō “entwine”, seru-ī, ser-tum;
    stert-ō “snore”, stertu-ī, ─ ;
    strep-ō “sound”, strepu-ī, ─ ;
    tex-ō “weave”, texu-ī, tex-tum;
    trem-ō “tremble”, tremu-ī, ─ ;
    uom-ō “vomit”, uomu-ī, ─ .

“the perfect in -uī had a great success: in the late adge, it replaced some old lengthened or sigmatic perfect: arduī, leguī, reguī instead of arsī¸ lēgī, rēxī” (Ernout, 1953, p. 208).

perfectum in [e:-wis]:

  • cern-ō “decree”, crē-u-ī, crē-tum;
    consuesc-ō “become accustomed“, consuēu-ī, consuē-tum;
    crēsc-ō “increase”, crē-u-ī, crē-tum;
    (ad)olēsc-ō “grow up”, adolē-u-ī, adultum;
    quiēsc-ō “rest”, quiē-u-ī, quiē-tum;
    ser-ō “sow”, sē-u-ī, sa-tum;
    spern-ō “scorn”, sprē-u-ī, sprē-tum;
    suēsc-ō “be wont”, suē-u-ī, suē-tus.
    pāsc-ō “feed”, pā-u-ī, pās-tum;
    stern-ō, strā-u-ī, strā-tum “strew”.
    cognosc-ō “get to know”, cognosce-re, cognō-u-ī, cogni-tum;
    ignosc-ō “forgive”, ignosce-re, ignō-u-ī, ignō-tum;
    nosc-ō “know”, nosce-re, nōu-ī, nō-tum.
  • with the same allomorph in perfectum as in infectum; and consequently with perfectum in /is/ or /…is/:
  • bibo “drink”, bib-ī (pō-tuī);
    īc-ō “hit”, īc-ī, ic-tum;
    -cend-ō “kindle”, -cend-ī,-censum;
    -scend-ō “climb”, -scend-ī, -scen-sum;
    cūd-ō “forge”, -cūd-ī, -cū-sum;
    -fend-ō “ward off”, -fend-ī, -fen-sum;
    mand-ō “chew”, mand-ī, man-sum;
    pand-ō “open”, pand-ī, pan-sum (pas-sum);
    pīns-ō (pīs-ō) “bruise”, pīns-ī, pīns-um (pīns-tum, pīs-tum);
    prehend-ō “seize”, prehend-ī, prehen-sum;
    scand-ō “climb”, ascend-ī, ascen-sum;
    sīd-ō “settle”, sīd-ī (-sēd-ī), -ses-sum;
    strīd-ō “whiz”, strīd-ī ─;
    uell-ō « pluck », uell-ī (-uul-sī), uul-sum;
    uerr-ō, uerr-ī, uer-sum “sweep”;
    uert-ō, uert-ī, uer-sum “turn”;
    animaduert-ō “pay attention to”, uert-ī, uer-sum;
    uīs-ō “go and look”, uīs-ī, uī-sum.
  • solu-ō “moose, pay”, solu-ī, solū-tum; [solw-o:] = /solu-o:/, [solw–i:] and [solu:-tus], phonetic realization of /soluu-t-us/;
    uolu-ō “turn”, uolu-ī, uolū-tum.
    consequ-or «follow », consequ-ī, consecū-t-us su-m .
  • with a reduplication perfectum:
  • cad-ō “fall”, ce-cid-ī, cā-sum;
    caed-ō “cut”, ce-cīd-ī, cae-sum;
    can-ō “sing”, ce-cin-ī, can-tum;
    scind-ō “tear”, scicid-ī (-scid-ī), sci-sum;
    tang-ō “touch”, tange-re, te-tig-ī, tac-tum;
    at-ting-ō “touch”, at-tinge-re, at-ø-tig-ī, at-tac-tum;
    cond-ō “put away” , conde-re, condid-ī, condi-tum ;
    pell-ō “push”, pelle-re, pe-pul-ī, pul-sum;
    perd-ō “ruin, destroy”, perdere, perdid-ī, perdi-tum ;
    posc-ō “ask”, posce-re, po-posc-ī ;
    prae-st-ō “be superior to others”, prae-stā-re, prae-sti-t-ī, prae-stā-tum (prae-sti-tum);
    prōd-ō “give birth to”, prōde-re, prōdid-ī, prōdi-tum;
    uend-ō “sell”, uende-re, uendid-ī, uendi-tum.
  • with a lengthening perfectum:
  • ag-ō “drive”, ēg-ī, āc-tum.
    ed-ō “eat”, ēd-ī, ē-sum;
    leg-ō “gather”, lēg-ī, lec-tum;
    em-ō “buy”, ēm-ī, emp-tum.
    cōg-ō “drive together” (=co+agō), cōge-re, coēg-ī, coac-tum;
    relinqu-ō “leave, abandon”, relinque -re, relīqu-ī, relic-tum;
    collig-ō “gather together, collect”, collige-re, collēg-ī, collec-tum.
  • 7.11.3. Some irregularities
  • • Verbs with consonantal alternation:
  • ger-ō “bear, carry”, gere-re, ges-s-ī, ges-tum (= /geri ~ ger/ and /ges/);
    quer-or “complain”, quere-ris, quer-ī, ques-t-us su-m (= /kweri ~ kwer/ and /kwes/ + Passive);
    ūr-ō “burn”, ūre-re, us-s-ī, us-tum (= /ūri ~ ūr/ and /us/);
    quaer-ō “seek”, quaesī-u-ī or quaesi-ī, quaesī-tum (= /kwairi ~ kwair/ and /kwaisi:/);
    hauri-ō “drain”, haurī-re “drain”, hau-s-ī, haus-tum (= /hauri:/ and /haus/).
  • nōsc-ō “know”, nosce-re, nōu-ī, nō-tum (= /nōsk/ and /nō/);
    cognosc-ō “get to know”, cognosce-re,cognō-u-ī, cogni-tum;
    ignosc-ō “forgive”, ignosce-re, ignō-u-ī, ignō-tum.
    crēsc-ō “increase”, crēsc-is, crē-u-ī, crē-tum (= /kre:sk/ and /kre:/);
    suēsc-ō “be wont”, suē-u-ī, suē-tus (= /suēsk/and /suē/);
    (ad)olēsc-ō “grow up”, adolē-u-ī, adultum;
    quiēsc-ō “rest”, quiē-u-ī, quiē-tum.
    pāsc-ō “feed”, pā-u-ī, pās-tum (= /pāsk/ and /pā/).
  • • Verbs with a suppletive stem:
  • toll-ō “pick up”, tolle-re, sustul-ī, sublā-tum.
    fer-ō “carry”, fer-s, tul-ī, lātum.
  • • Defective verbs, the signifier of which is always constituted by a morphological unit of Perfectum:
  • coep-ī “I began”, coep-is-tī, coep-is-se;
    memin-ī “I remember”, memin-is-tī, memin-is-se;
    ōd-ī “I hate”, ōd-is-tī, ōd-is-se.
  • • Semi-deponent verbs, which always add a Passive morphological unit to the Perfectum morpheme:
  • aude-ō “dare”, au-s-us su-m;
    gaude-ō “rejoice”, gāuī-s-us su-m ;
    sole-ō “be wont”, soli-t-us su-m;
    fīd-ō “trust”, fīde-re, fī-s-us sum.
  • • Verbs with active and deponent forms fairly equally distributed:
  • mere-ō “I deserve” , merē-re, mer-u-ī, meri-tum or mere-or “I deserve”, merē-rī, meri-t-us su-m «mériter».

Retour au plan ou Aller au § 8.

1) Cf. Touratier, 1994, p. 94-101.
2) Priscien, VIII, 93-94, K II, p. 442.28-443.10: cum enim omnia uerba, quae aequali regula declinantur, in o uel in or desinant, in o quidem terminantia, si primae sint coniugationis, in as efferunt secundam personam, ut oro oras, sto stas; sin secundae, in es, ut moneo mones, haereo haeres; sin tertiae, in is correptam, ut cupio cupǐs, curro currǐs; sin quartae, in is productam, ut munio munīs, esurio esurīs. in or uero uerba primae coniugationis in aris faciunt secundam personam, ut amor amaris, criminor criminaris, luctor luctaris; secundae in ēris producta paenultima, ut doceor docēris, reor rēris; tertiae in ĕris paenultima correpta, ut legor legĕris, sequor sequĕris; quartae in iris, ut munior muniris, molior moliris, audior audiris
3) Donatus, Ars grammatica II 12, Keil IV p. 381 0591: Coniugationes verborum quot sunt? Tres. Quae? Prima secunda tertia. Prima quae est? Quae indicativo modo tempore praesenti numero singulari secunda persona verbo activo et neutrali a productam habet ante novissimam litteram, passivo communi et deponenti ante novissimam syllabam, ut amo amas, amor amaris; et futurum tempus eiusdem modi in bo et in bor syllabam mittit, ut amo amabo, amor amabor. Secunda quae est? Quae indicativo modo tempore praesenti numero singulari secunda persona verbo activo et neutrali e productam habet ante novissimam litteram, passivo communi et deponenti ante novissimam syllabam, ut doceo doces, doceor doceris; et futurum tempus eiusdem modi in bo et in bor syllabam mittit, ut doceo docebo, doceor docebor. Tertia quae est? Quae indicativo modo tempore praesenti numero singulari secunda persona verbo activo et neutrali i correptam vel i productam habet ante novissimam litteram, passivo communi et deponenti pro i littera e correptam vel i productam habet ante novissimam syllabam, ut lego legis, legor legeris, audio audis, audior audiris; et futurum tempus eiusdem modi in am et in ar syllabam mittit, ut lego legam, legor legar, audio audiam, audior audiar. haec in imperativo et in infinitivo statim discerni possunt, utrum i littera correpta sit an producta. .
4) Cf. Cic., orator157 : Nec uero reprenderim : « scripsere alii rem » ; « scripserunt » esse uerius sentio ; sed consuetudini auribus indulgenti libenter obsequor
5) James Foley, 1967, La prothèse dans le verbe latin sum , in : Langages 8, dec. 1967, p. 60-66.
6) This synchronic explanation makes it possible to resolve a diachronic problem of the italic comparative linguistics. “The Osq form fufans < *bhu-bha-nt ‘erant’, the only imperfect example out of Latin in the Italic languages” (according to Monteil, 1970, p. 327), confirm the italic origin of the -bā- Imperfect morpheme. But it is surprising that Latin has only -ā- in a verb where Osq has the equivalent to the Latin -bā-. A historic assumption similar to that of J. Foley would be better than resorting to a simple and vague “analogical extension peculiar to Latin” (Monteil, p. 327
7) The sp, st, sc groups are the only interior groups in s that are possible in Latin, as asper, festus, fascis (cf. Touratier, 2005, Système des consonnes, p. 124 n. 5).
8) Cf. Touratier, 2005, p. 1278-128.
9) Sus-tul-ī and sublā-t-um also supply the perfect and supine of the verb toll-ō “pick up”.
10) Plaut, Rud. 1242: Mihi istaec uidetur praeda praedatum irier “this appears to be plunder that will soon be plundered from you again” (Cleveland K. Chase, 1919
11) Plaut., Pers. 577-8 : Veniri hanc uolo, Si potest “I want her to be sold, if possible”.
12) cf. Niedermann, Max, 1953 , Phon. historique, p. 69.
13) Cf. Niedemann, 19533, p. 121.
14) Cf Niedermann, 1953, p. 149.
15) Cf. Niedermann, 19533, p. 148
16) Cf. Niedermann, 1953, p. 148
17) Ederent is nevertheless found in Gel. 19,2,7.
18) Plaut, Men. 147 : Vbi essuri sumus ? “Where are we going to eat?”