7. Verbs

7.1. Introduction

7.1.1. Voice, Mood, Tense, Person, Number

1. The inflection of the verb is called its conjugation.

2. Through its conjugation the verb expresses voice, mood, tense, person and number.

3. The voices are two: active and middle (or more exactly middle-passive).

4. The moods were up to five:

a. The indicative, for plain statement of objective fact.

b. The imperative, for commands.

c. The optative, for intentions or hopes for action.

d. The subjunctive, for potentiality, possibility.

NOTE. An old injunctive is also reconstructed to account for S.LIE forms; v.i. §7.4.2.

5. The general tenses are three, viz.:

a. The present.

b. The past.

c. The future.

NOTE. The so-called future stem is generally believed to have been an innovation in post-LIE, not spreading to some dialects before the general split of the proto-languages; however, the distinction between a present and a future tense is common to all IE languages.

6. The aspects were up to three:

a. For continued, not completed action, the present.

b. For the state derived from the action, the perfect (or more exactly stative).

c. For completed action, the aorist.

7. There are three verbal tense-stems we will deal with in this grammar:

I. The present stem, which gives the present with primary endings and the imperfect with secondary endings.

II. The aorist stem with secondary endings, giving the aorist (always past), usually in zero grade, with dialectal augment.

III. The perfect stem, giving the perfect, only later specialised in present and past.

NOTE. From this reconstructed original PIE verbal system, a future stem was created from some present stem formations. The aorist merged with the imperfect stem in Northwestern dialects, and further with the perfect stem in Germanic, Italic, Celtic and Tocharian. The aorist, meaning the completed action, is then reconstructed as a third PIE aspect, following mainly the findings of Old Indian, Greek, and also – mixed with the imperfect and perfect stems – Latin.

8. The persons are three: first, second, and third.

9. The numbers in the Indo-European verb are two: singular and plural, and it is the only common class with the noun. It is marked very differently, though.

NOTE. The reconstructed dual in the verbal system seems to have been a late development, systematised only after the LIE split in some dialects, and disappeared in others, see §3.5.

7.1.2. Voice

1. In grammar, voice is the relationship between the action or state expressed by a verb and its arguments.

2. When the subject is the agent or actor of the verb, the verb is said to be in the active. When the subject is in some way affected by the verbal action, it is said to be in the middle-passive.

NOTE. For Clackson (2007): “The middle is the voice used to denote that the subject is in some way affected by the verbal action. Thus, for transitive verbs the active typically represents the subject as the actor, and the middle represents the subject as the undergoer. For intransitive verbs the middle is preferred when there is some notion of control over the verbal action (hence the middle inflection of ‘think’ and ‘speak’), but if the verb denotes an event or action where the participant cannot have control, the active is used (thus ‘be’, ‘vomit’ and ‘wait’).”

2. The active and middle-passive voices in Indo-European generally correspond to the active and passive in English. The middle had these uses in LIE (Clackson 2007):

a. Personal involvement, or sense of benefaction for the subject; as, Skt. yájati (active) ‘(s)he performs a sacrifice’ (said from the priest), and  yájate ‘(s)he performs a sacrifice’ (said of person for whose benefit the sacrifice is made).

b. Reflexivity, generally referring to an action whose object is the subject, or an action in which the subject has an interest or a special participation; as, Greek lówō (active), I wash, lówomar (middle), I wash myself, or wéstijetor, (s)he dresses (him/herself), (s)he gets dressed.

c. Reciprocity; as, Hitt. appanzi (active) ‘they take’, Hittite SU-za appantat (lit. hand take-middle) ‘they took each other by the hand’.

b. Passivity, usually meaning that an action is done. This is the default meaning in Italo-Celtic and Germanic, also found in the Greek and Anatolian middle; as, stoighōs péwontor, streets are (being) cleaned.

NOTE. According to Fortson (2004): “The middle could also express the passive voice, which indicates that the subject is acted upon by someone else: ‘is being fought’, ‘was washed’. A tradition of scholarship rejects positing a passive voice for PIE because there was no separate set of passive endings. But all the daughter languages that have a separate passive conjugation have developed it in whole or in part from the PIE middle endings, and it seems best to regard the middle as having been, in fact, a mediopassive or middle-passive – capable of expressing either voice depending on the context.”

3. Apart from this middle-passive voice system, the relic of an old impersonal -()r desinence is reconstructed for LIE times.

NOTE. This desinence begins in the third person (singular or plural), according to Jasanoff (“The r-endings of the Indo-European middle”, Sprache 23, 1977) and Szeméreny (1985), and probably not as middle mark, as proposed by Kuryłovicz (Indogermanische Grammatik II Akzent-Ablaut, 1968) and Bader (“Relations de structure entre les desinences d’infectum et de perfectum en latin”, Word 24, 1968); it is indeed probably at the origin of middle-passive primary endings in -r, though, but was apparently used as impersonal mark within the active voice. It is the oldest reconstructed meaning of the -r ending, and it is only found originally in the 3rd sg. and 3rd pl. (probably originally without lengthening or distinction of singular vs. plural). The ending is reconstructed as *-()r. It is found in Italo-Celtic, Germanic (cf. O.H.G. skritun), Tocharian, Old Indian, and possibly in Armenian. Adrados–Bernabé–Mendoza (1995-1998).

It marked impersonality, as in Celtic impersonal verbal forms; it is similar to Eng. ‘there’ in ‘there are three books’, and equivalent to Ger. ‘es wird’ in ‘es wird geschlafen’, or Spa. ‘se’ in ‘aquí se duerme’. It was attached directly to the present, aorist or perfect stem. So e.g. kei sweper, one sleeps here, ‘it is slept here’, edjēu wgjer, today one works, ‘it is worked today’.

4. Some verbs are only active; as, esmi, be, edmi, eat, or dōmi, give.

5. Many verbs are middle-passive in form, but active or reflexive in meaning. These are called deponents in Latin; as, gnāskar, be born, keimar, lie, lay; séqomar, follow, etc.

7.1.3. Moods

1. The mood in which a verb appears expresses the speaker’s attitude or stance taken towards the action – whether (s)he is asserting that it is factual, or indicating a wish that it were or were not true, or reporting the action second-hand, or indicating a contrafactual condition (Fortson 2004).

NOTE. While the oldest PIE had possibly only indicative and imperative, a subjunctive and an optative were common in Late Indo-European, both used in the present, perfect and aorist. Not all dialects, however, developed those new formations further into a full system.

2. The indicative mood is used for most direct assertions and interrogations, with the action described as fact.

3. The imperative is used for exhortation, entreaty, or command.

3. The subjunctive mood is used for actions described as completely theoretical, generally with a future meaning, frequently translated by means of the auxiliaries may, might, would, should.

c. The optative is used to express wishes, hopes, and various other non-factual modalities.

7.1.4. Aspect

1. Aspect is a grammatical category that refers to the type of action indicated by a verb. Actions can be done once or repeatedly, to completion or not, or be ongoing with neither a true beginning nor end.

2. According to the generally accepted view, the imperfect and aorist were distinct aspectually, the imperfect expressing incomplete or ongoing action in past time (imperfective aspect), the aorist indicating completed or punctual (one-time) action in past time (perfective aspect).

3. The perfect or stative referred to a current state derived from the action (resultative aspect), without a temporal reference.

NOTE. A number of archaic examples of the perfect, especially in Indo-Iranian and Greek, refer to states in present time; so e.g. Lat. meminit ‘remembers’, Gk. mémone, Goth. man ‘is mindful of’, ‘thinks’. Furthermore, the singular perfect endings are sued to inflect a class of presents in Anatolian, the so-called i-conjugation. These facts together have led researchers to believe that the PIE perfect was a stative. In the dialects attested, however, except for relic forms like these, perfect forms express past tense; to explain this development, it is usually said that the PIE stative perfect had (or optionally had) resultative overtones (‘is in a state resulting from having done X’, therefore ‘has done X’). Fortson (2004).

The Late Indo-European aspectual system might be so depicted, using a amore exact vs. the  traditional notation:

Process

Aspect

Aspect (traditional)

Tense

Tense (traditional)

Stative

Stative

Perfect system

(unmarked)

Perfect tense

Eventive

Perfective

Aorist system

(unmarked)

Aorist tense

Imperfective

Present system

Present

Present tense

Past

Imperfect tense

The terminology around the stative, perfective and imperfective aspects can be confusing. The use of these terms in this table is based on the reconstructed meanings of the corresponding forms in PIE and the terms used broadly in linguistics to refer to aspects with these meanings. In traditional PIE terminology, the forms described in the above table as stative, perfective and imperfective are known as the perfect, aorist and present systems. The present/imperfective system in turn can be conjugated in two tenses, described here as present and past but traditionally known as present and imperfect. The traditional terms are based on the names of the corresponding forms in Ancient Greek (also applied to Sanskrit), and are commonly encountered. The existence of the terms ‘perfect’ and ‘perfective’, with quite different semantics, is especially problematic, and is a perennial source of confusion in linguistics as a whole.

7.1.5. Tenses of the Finite Verb

The tenses of the indicative have, in general, the same meaning as the corresponding tenses in English:

a. Of continued action

I. Present: deikō, I show, I am showing, I do show

II. Imperfect: deikom, I was showing, I used to show

b. Of completed action or the state derived from the action

V. Aorist: dikóm, I showed, I had shown, I have shown

c. Of state derived from the action

IV. Perfect: doika, I am in a state derived from having shown.

7.2. Forms of the Verb

7.2.1. The Verbal Stems

1. The actual forms of the tenses and moods were made from tense-stems, of which there were up to four for each verb, called (1) the present, (2) the aorist, (3) the perfect; and also (4) future stems in post-LIE times.

NOTE. There are some characteristic forms of each stem, like the suffix -n- or -sko, which give mostly present stems. Generally, though, forms give different stems only when opposed to others.

2. The different stems are used in the verbal conjugation as follows:

STEMS

WHERE USED

Present

Present and Imperfect (active and middle-passive)

Aorist

Aorist (active and middle-passive)

Perfect

Perfect

Future

Future

3. There are some monothematic verbs, as esmi, be, or edmi, eat – apparently remains of the oldest PIE. And there are also some traces of recent or even nonexistent mood oppositions. To obtain this opposition there are not only reduplications, lengthenings and alternations, but also ablaut and accent shifts.

NOTE. Not every verb could form all three tense-stems. Quite a few did not form perfects, for example, and derived verbs only had present stems in PIE. IEDs did usually however innovate additional tense-stems for these verbs.

4. Most Late Indo-European verbs are built with a series of derivational suffixes that alter the root meaning, creating denominatives and deverbatives. The first are derived from nouns and adjectives; as, torsējō, dry, “make dry”, from ters-, dry, or newājō, make new, from new-, new.  The last are derived from verbs, as widējō, see, from weid-, see, know.

NOTE. It is not clear whether these deverbatives – causatives, desideratives, intensives, iteratives, etc. – are actually derivatives of older PIE roots, or are frozen remains, formed by compounds of older PIE independent verbs added to other verbs, the ones regarded as basic.

5. Reduplication is another common resource; it consists of the repetition of the root, either complete or abbreviated; as, sisdō, sit down, settle down, from sed-, sit, gígnōskō, know, from gnō-, mímnāskō, remember, from men-, think, etc.

6. Thematic e/o has no meaning in itself, but it helps to build different stems opposed to athematics. Thus, It can be used to oppose a) indicative athematic to subjunctive thematic, b) present thematic to imperfect athematic, c) active to middle voice, etc. Sometimes accent shift helps to create a distinctive meaning, too.

7. Stems are inflected, as in the declension of nouns, with the help of vowel grade and endings or desinences.

7.2.2. Verb-Endings

1. Every form of the finite verb is made up of two parts:

I. The stem. This is the root or an extension, modification or development of it.

II. The ending or desinence, consisting of:

a. The signs of mood and tense.

b. The personal ending.

So e.g. the root deik-, show, lengthened as thematic present verb-stem deik-e/o-, to show, and by the addition of the personal primary ending -ti, becomes the meaningful déik-e-ti, he shows.

2. Verbal endings can thus define the verb stem, tense and mood. Fortson (2004):

DESINENCES

WHERE USED

Primary active                       

present indicative active, active subjunctives

Secondary active

imperfect and aorist indicative active, active optatives

Primary middle-passive

present indicative middle, middle subjunctives

Secondary middle-passive

imperfect and aorist indicative middle, middle optatives

Perfect

perfect

3. The primary series indicates present and future; sg. -mi, -si, -ti, and 3rd pl. -nti are the most easily reconstructed LIE formations. The secondary endings indicate past; sg. -m, -s, -t and 3rd pl. -nt. The subjunctive is marked with primary desinences, while the optative is usually marked with secondary endings. The imperative has or special endings.

The secondary endings are actually a negative term opposed to the primary ones. They may be opposed to the present indicative, they may indicate indifference to tense, and they might also be used in the present.

They can also mark the person; those above mark the first, second and third person singular and third plural.

Also, with thematic vowels, they mark the voice: -ti active primary | -t active secondary; -tor/-toi middle primary | -to middle secondary.

4. The augment appears in Ind.-Ira., Gk., and Arm., to mark the past tense (i.e., the aorist and the imperfect). It was placed before the stem, and consisted generally of a stressed é-, which is a dialectal Graeco-Aryan feature not found in N.LIE.

Active Endings

1. The characteristic active primary endings are singular -mi, -si, -ti, 3rd plural -nti, while the secondary don’t have the final -i, i.e. sg. -m, -s, -t, 3rd pl. -nt.

NOTE. The secondary endings are believed to be older, being originally the only verbal endings available. With the addition of a deictic -i, termed the ‘hic et nunc’ particle (Latin for ‘here and now’), the older endings became secondary, and the newer formations became the primary endings. It may have been the same as the -i found in pronominal and adverbial forms. Adrados–Bernabé–Mendoza (1995-1998), Fortson (2004). Compare a similar evolution in Romance languages from Lat. habere, giving common Fr. il y athere (it) is’, or Cat. i hathere is’, while the Spanish language has lost the relationship with such older Lat. I ‘there’, viz. Spa. hay ‘there is’ (from O.Spa. ha+i), already integrated within the regular verbal conjugation of the verb haber.

2. These desinences are used for all verbs, whether athematic or thematic; as, esti, he is, or déiketi, he shows. However, in the 1st sg., thematics end in -ō; as, deikō.

NOTE. These endings in -ō are probably remains of an older situation, in which no ending was necessary to mark the 1st sg. (that of the speaker), and therefore, even though a desinence -m became general with time, the older formations prevailed, along with a newer thematic -o-mi.

3. The thematic and athematic endings of the active voice are reconstructed as follows:

 

 

Athematic

Thematic

Sg.

 

Primary

Secondary

Primary

Secondary

Sg.

1.

-mi

-m

-ō

-om

2.

-si

-s

-esi

-es

3.

-ti

-t

-eti

-et

Pl.

1.

-mes

-me

-omos

-omo

2.

-te

-te

-ete

-ete

3.

-ti

-t

-onti

-ont

NOTE.  About the reconstruction of PIE active endings:

1) 1st p. pl. thematic endings -o-mo, -o-mos, are found in Italic (Lat. -mus), Celtic (O.Ir. *-mo or *-mos), Balto-Slavic (cf. Pruss. -mai, O.C.S. -mŭ<*-mo, *-mos or *-mom), and from -mo- or -me-,  in Germanic (cf. Goth. -m) and Indo-Iranian (cf. O.Ind. -ma). Primary Thematic ending -o-mo- does not have a clear ‘original’ PIE ending, but an -s the most logical choice, given the contrast between primary endings -mes/-me. So in Ringe (2006). Beekes (1995) tentatively reconstructs  -o-mom-.

2) 2nd P. Pl. ending athematic -the (<*-tHé) is only found differentiated in PII, while in the other dialects it would have evolved as a common -te.

3) Athematic desinences in *-enti, as found in Mycenaean and usually reconstructed as proper PIE endings, weren’t probably common PIE desinences. Compare Att. Gk. -aasi (<-ansi<-anti), or O.Ind. -ati, both remade from an original zero-grade PIE *-n̥ti. In fact, Mycenaean shows some clearly remade examples, as Myc. e-e-esi<*esenti (cf. Ion. εων), or ki-ti-je-si (<ktíensi).

Middle-Passive and Perfect Endings

1. The middle-passive endings are generally those of the active voice with a characteristic middle voice -o, in which the primary endings have an additional -i or -r, depending on the dialects.

NOTE. In the moods, the endings attested are the same. Only dialectally were some new endings developed to differentiate the subjunctive.

2. There were apparently two possible set of endings already in Late Indo-European: either because the original primary -r endings were replaced by the endings in -i, or because both came to be used with the same meaning by different dialects at the same time, with an expansion of that use through neighbouring contact zones. Syncretic trends led in any case to dialectal specialisation of both marks into the known (middle, middle-passive or passive) systems attested.

NOTE. Italic, Celtic, Tocharian, and Phrygian had mediopassive primary endings in -r (cf. Lat. -tur, O.Ir. -tha(i)r, Toch. -tär, Phryg. -tor), whilst others show -i (cf. Skt., Av. -te, Gk., Toch. -tai, Goth. -da); both forms coexisted in Anatolian (with -r as primary ending, combined with -i cf. Hitt. -ta-r-i, nta-r-i), in Tocharian (with -r as primary ending, -i as secondary), Indo-Iranian (with -r- passive forms), and afield cf. also Germanic (with remains of forms in -r with impersonal value, cf. O.H.G. skritun).

From these finds it is thought that -r was the old ‘original’ primary middle marker (possibly taken from the impersonal mark, v.s. §7.1.2), corresponding to the -i of the active. Both mediopassive endings (-r and -i) coexisted already in the earliest reconstructible PIE, and -i probably began to replace the old impersonal -r as the general middle marker already by Late Indo-European, as the Anatolian endings -r- and Sanskrit remains of middle forms in -ro- show. In the northern dialects -r became later specialised for the passive constructions or disappeared. It is therefore reasonable to think that while the general trend in N.LIE was to keep (or generalise) the middle-passive with primary endings in -r, in S.LIE the middle-passive in -i gradually replaced the older endings. Endings in -rgeneralised in Phrygian and impersonal -r in Armenian (both S.LIE dialects), and endings in -i generalised in Germanic (a N.LIE dialect), which also shows traces of an old -r, further complicate the situation, showing that specialisation trends – at least in a post-IED period – were not uniform.

Fortson (2004): “This *-r is now generally thought to have been the primary middle marker, corresponding to the *-i of the active. Middles in Anatolian, Italic, Celtic, Tocharian, and Phrygian preserve this -r, but it has been replaced by the -i of the active in Indo-Iranian, Greek, Germanic, and Albanian”.

Clackson (2007): “The morph *-r appears therefore to have acted as the analogue to *-i in the active endings and originally marks the ‘here and now’ of middles. This explanation, proposed by Cowgill (1968:25-7), also accounts for the absence of *-r in Greek and Vedic middle endings: at one stage these languages must have replaced the primary marker *-r with *-i on the analogy of the active endings.”

Ringe (2006) also considers the hic-et-nunc particle of the mediopassive to have been *-r rather than *-i, having been replaced by the spread of active *-i.

Beekes (1995) considers the addition of *-r or *-i to the ‘original’ (secondary) endings to be a late development from a PIH point of view: “From this it follows that the -r was not characteristic of the primary endings. But neither was the -i of Sanskrit and Greek the marker of the primary ending in PIE, because the languages which generalized the -r show no trace of the -i. That on the other hand especially the marker of the primary endings of the active (the -i) in some languages was also used for the middle, is understandable. The conclusion is that there was no opposition between primary and secondary (…) we shall see that the perfect endings can be considered as secondary endings of the middle (…)”.

Adrados–Bernabé–Mendoza (1995-1998) also deem both marks *-r and *-i to have coexisted parallel to each other as systems to mark the primary endings, as a late development.

3. The thematic and athematic endings of the middle-passive, reconstructed from a Northern Late Indo-European point of view, are as follows:

MIDDLE-PASSIVE AND PERFECT ENDINGS

 

ATHEMATIC

THEMATIC

STATIVE

 

Primary

Secondary

Primary

Secondary

Sg.

-mar

-ma

-ar

-a

-a

-sor

-so

-esor

-eso

-tha

-tor

-to

-etor

-eto

-e

Pl.

-mesdha

-medha

-omesdha

-omedha

-

-(s)dhwe

-dhwe

-e(s)dhwe

-odhwe

-

-tor

-to

-ontor

-onto

-()r

NOTE. 1) The Southern Late Indo-European middle primary endings in -i are easily reconstructed for the singular, -ai, -soi, -toi, and the 3rd plural -ntoi, even though Toch. B (secondary endings) -tai, -te, -nte still suggest to some (Neu 1968) that the original PIE were *-sai, *-tai, *-ntai, instead of the general opinion. Dialectal Greek forms in the singular point to an alternative 1st sg. -oi. A full discussion of the reconstruction is found e.g. in Villanueva Svensson’s <http://eprints.ucm.es/tesis/fll/ucm-t26697.pdf>.

2) The forms of the 1st and 2nd person plural were apparently the same in both systems. Greek, Indo-Iranian, and Anatolian dialects show middle-passive 2nd plural forms in -medha(<*-medh-h2, O.Ind. -mahe, Gk. -metha, Toch. -ämtä-), -mesdha(<*-mesdh-h2, cf. Gk. -mestha, Hitt. -wašta-), Toch. -cär (cf. Toch. -t<-dhwe), O.Ind. mahi>*-megha, and -men, cf. Gk. -men, Hitt. -wen-i

1st pl. *-mo(s)r, Lat. -mur, and 2nd P. Pl.  Osc. -ter, Hitt. -ttumari,. In Ita.-Cel. a form *-ntro has been related to the perfect, hence related to the ‘original’ paradigm with an 3rd p. pl. secondary -ro, primary -ro-r?; a 3rd p.sg. -o, -or is also reconstructible.

4) The forms in -r are reconstructed according to Kortlandt (1979), Sihler (1995), Beekes (1995), Fortson (2004), and Clackson (2007); all of them make a similar account of the ‘older’ paradigm, which includes the forms 3rd sg. prim. -or, sec. -o, 3rd pl. prim. -ro, sec. -ront, attested in scattered remains in Hittite, Sanskrit, Tocharian, Sabellian, and Old Irish, what suggest that they were the ‘original’ ones, being replaced by the common endings. The old middle-passive ending system was then apparently sg. -a-, -tha-, -o-/-to-, pl. -ro-/-nto-, to which primary endings were attached in -i, *-so-i, *-to-i , *-nto-i, or in -r, *-ar, *-tar, *-or, pl. *-ro-r?/*-ntor, from older *-h2-, *-th2-, *-o, pl. *-r.

These endings share similarities with the perfect ones, cf. *-h2e, *-th2e-, *-e, pl. *--, *-é,  *-r. About the different writing of *-th2e-, as -tha or -ta, it is not only restricted to Proto-Indo-Iranian; cf. perf. Gk. oĩstha.

This similarity of perfect and ‘original’ middle endings is explained differently according to the available theories on the prehistory of PIE verb (through internal reconstruction), apparently involving complicated syncretic and innovative trends regarding the voices, tenses and aspects. However, what seems clear from the later developments attested in the older IE languages, is that the synchronic picture of the Late Indo-European middle and perfect verbal ending system had to be near to the one depicted above.

Dual Endings

A complete reconstruction of the dual endings is not possible, because there is too little and contradictory data, probably because of the late development of the verbal dual (see above §3.5).

Only the active paradigm shows common endings:

 

 

Primary

Secondary

Du.

1.

-wes

-we

2.

-t(h)os

-tom

3.

-tes

-tām

NOTE. Dual endings are found in Ind.-Ira., Gk., BSl. and Gmc., but apart from a common 3rd prim. -tom / sec. -tām in O.Ind. and Gk., there is only a general (usually incomplete) paradigm 1st w-, 2nd & 3rd t-, with different lengthenings in *-e/-o, *-es/-os, *-ā. This table has ben taken from Beekes (1995). Fortson (2004) reconstructs an uncertain -to- for 2nd and 3rd. Only Beekes tentatively reconstructs uncertain middle endings for the parent language.

7.2.3. The Thematic Vowel

1. Stem vowels are – as in nouns – the vowel endings of the stem, especially when they are derivatives. They may be i, u, ā, ē (and also ō in roots). But the most extended stem vowel is e/o (also lengthened ē/ō), called thematic vowel (see above §2.6) which existed in PIH before the split of the Anatolian dialects, and which had overshadowed the (older) athematic stems already by Late Indo-European. The thematisation of stems, so to speak, relegated the athematic forms especially to the aorist and to the perfect; many old athematics, even those in -ā- and -ē-, are usually found extended with thematic endings -je/o-.

NOTE. The old thematics were usually remade, but there are some which resisted this trend; as edmi, I eat, dōti, he gives, or idhi! go!

The stem vowel has sometimes a meaning, as with -ē- and -ā-, which can indicate state. There are also some old specialisations of meanings, based on oppositions:

a. Thematic vs. athematic:

- Athematic indicative vs. thematic subjunctive. The contrary is rare.

- Thematic present vs. athematic aorist, and vice versa.

- It may also be found in the middle-active voice opposition.

b. Thematic stem with variants:

- The first person, thematic in lengthened -ō.

- Thematic o in 1st sg. & pl. and 3rd pl.; e in 2nd and 3rd sg. and 2nd pl. There are also archaic 3rd pl. in e, as senti, they are.

c. Opposition of thematic stems. This is obtained with different vowel grades of the root and by the accent position.

2. In the so-called semithematic inflection, ahematic forms alternate with thematic ones.

NOTE. The semithematic inflection is for some an innovation of LIE, which didn’t reach some of the dialects, while for other scholars it represents a situation in which the opposition thematic-athematic and the accent shifts of an older PIE system had been forgotten, leaving only some mixed remains within an already generalised LIE regular thematic verbal system.

7.3. The Conjugations

7.3.1. Conjugation is the traditional name of a group of verbs that share a similar conjugation pattern in a particular language, a verb class. Late Indo-European shows regular conjugations, and all verbs may be conjugated in any person, number, tense, mood and voice by knowing which conjugation it belongs to.

NOTE. According to Clackson (2007): “In order to explain the number of different stem formations with the same function, Meillet supposed that in the parent language not just one present stem was opposed to one aorist stem, but rather it was possible to form several present and aorist stems from the same root. These stems were held to show different ‘nuances’ of aspectual meaning (or, to use the German term, Aktionsart), such as punctual, repeated or incipient action. Each root could show a wide variety of different formations, none of which presupposed the other. (…)

However, better knowledge of the earliest attested IE languages has led to a revision of this view, and researchers have increasingly become aware that if two stems can be reconstructed for PIE, one may represent an archaism and the other an innovatory replacement. Thus athematic verbs are in general a relic class, replaced over the history of individual languages by thematic formations. Motivation for the replacement of athematic verbs is not difficult to find: the juxtaposition of root-final consonants and the athematic endings (mostly consonant-initial) led to clusters which were often simplified or otherwise altered, so that the boundary between root and desinence, or suffix and desinence, became opaque to speakers. In some languages, paradigms still survive which exemplify the extent to which regular phonological developments can conceal the form of the root and the suffix. (…)”

7.3.2. A reference classification of PIE verbs into conjugations is the Lexikon der indogermanischen Verben (2001), supervised by H. Rix. It offers a conservative approach, not including much information on the Hittite verb (and thus Anatolian), being thus very interesting as a real approach to a living Late Indo-European verb.

NOTE. Nevertheless, it features a PIH phonetic reconstruction, and looks for the ‘original’ pre-LIE lexicon, what makes the work fit somewhere between the conventional PIH reconstruction and the modern PIH/LIE distinction, being thus somehow inconsistent, and at best showing a picture of a pre-LIE verbal system. It is therefore complex for a simple grammar, and indeed not directly applicable to an IED scheme, in which some athematic paradigms had been lost (or frozen into scarce, hence irregular examples), while newer verbs (and remade ones) further split within the most productive classes. Nevertheless, it is no doubt the most comprehensive work to date on the Proto-Indo-European verb.

7.3.3. The most important LIV verbal classes is as follows (Meier-Brügger 2003):

LIV

STEM CLASS

Examples

1a

Present, athematic, amphidinamic root

*gwhen-ti/*gwhn-énti alternance with -o- + -Ø- = -o-

1b

Present, athematic, acrodynamic root

*stēu-ti/*stéw-n̥ti n̥ti.alternance with -o- + - o - = -ō-

1g

Present, athematic, with -e- reduplication

*dhé-dhoh1-ti/*dhé-dhh1-n̥ti

1h

Present, athematic, with -i- reduplication

*sti-stéh2-ti/*sti-sth2-énti

1i

Present, thematic, with -i- reduplication

*gi-gn̥h1-é-ti

1k

Present, athematic, with nasal Infix

*li-né-kw-ti/li-n-kw-énti

1n

Present, thematic suffix -e-, e grade root

*bhér-e- ti

1o

Present, thematic suffix -é-, zero grade root

*ghr̥h3-é- ti

1p

Present, thematic suffix -ské-, zero grade  root

*gwm̥-ské- ti

1q

Present, thematic suffix -jé-, zero grade root

*gn̥h1-jé-toi

2a

Aorist, athematic, root

*gwem-t

2b

Aorist, athematic, suffix -s-

*prek-s-n̥t

2c

Aorist, thematic, reduplicated

*we-ukw-e-t

3a

Perfect, reduplicated

*gwe-gwom-/gwe-gwm-

7.3.4. We have divided the Late Indo-European verbs in two main conjugation groups: athematic and thematic. The latter were the most productive and abundant ones in IEDs, and often replaced the older athematics by means of derivation.

Athematic and thematic groups are, in turn, subdivided into four and eight subgroups respectively.

 

A. The Thematic Conjugation

The thematic conjugation group is formed by the following 8 subgroups:

I. Root verbs with root vowel e in the present and o in the perfect:

a.    Triliteral: deikō, dikóm, doika, deiksō, show, etc.

b.   Concave: teqō, teqóm, toqa/tōqa, teqsō, escape, séqomai, follow, etc.

NOTE. For IE teqō, cf. O.Ir. téchid/táich (<e/ō).

II. Concave root verbs with non-regular perfect vocalism. Different variants include:

a.    labhō, lābha, take; lawō, lāwa, enjoy, slabai, slāboma, fall (Middle Voice); aisdai, praise.

NOTE. Compare Gk. αιδομαι, O.ind. ile, Gmc. part. idja-.

b.   kano, kékana/kékāna, sing.

c.    legō, lēga, join, read, decide.

d.   lowō, lōwa, wash.

e.    rādō, rāda, shuffle, scrape, scratch.

f.     rēpō, rēpa, grab, rip out.

g.    rōdō, rōda, gnaw.


 

III. Verba vocalia, i.e., thematic --je/o-, --je/o-, -í-je/o-, -ú-je/o-:

a.    amājō, love.

e.    lubhējō, love, desire.

i.     sāgijō, look for, search.

u.   argujō reason, argue (cf. Lat. arguō, Hitt. arkuwwai).

o.   Causative-iteratives in -ejo-: bhoudhejō, wake somebody up.

IV. Verbs in -je/o-:

a.    Triliteral:  kupjō, kup(j)óm, koupa, keupsō, demand, desire, tremble.

b.   Concave: jakjō, jēka, throw.

c.    Lamed-he: parjō, pepra/péprōka, produce.

d.   Reduplicated Intensives: kárkarjō, proclaim, announce (cf. Gk. καρκαίρω, but Skt. carkarti).

V. Intensives-inchoatives in -ske/o-:

a.    Of mobile suffix: swēdhskō, swēdhjóm, swēdhwa, swēdhsō, get used to.

b.   Of permanent suffix: pksk, inquire.

VI. With nasal infix or suffix:

a.    Perfect with o vocalism: jungō, jugóm, jouga, jeugsō, join.

b.   Reduplicated perfect: tundō, tét(o)uda/tút(o)uda, strike.

c.    Convex: bhrangō, bhrēga, break.

d.   Nasal infix and perfect with o root: gusnō, gousa (cf. Lat. dēgūnō, dēgustus)

e.    Nasal infix and reduplicated perfect: cf. Lat. tollō, sustulii (supsi+tét-), lift.

VII. With reduplicated present:

a.    sisō, sēwa, sow.

b.   gignō, gegna, gégnāka, produce.

VIII. Other thematics:

o     p, pép(o)la.

o     widē, woida, see.

o     etc.

B. The Athematic Conjugation

Verbs of the second or athematic conjugation group may be subdivided into: 

I. Monosyllabic:

a.    In consonant: esmi, be, edmi, eat, ēsmai, find oneself, be.

b.   In ā (<*-h2): snāmi, swim, bhāmai, speak.

c.    In ē (<*-h1): bhlēmi, cry, (s)rēmai, calculate.

d.   With nasal infix: leiq- (lineqti/linqti), leave, kleu- (kneuti/knunti), hear, peu- (punāti/punānti), purify, etc.

NOTE. These verbal types appear mostly in Indo-Iranian and Hittite examples, and could therefore be more properly included in the suffixed (BIVc) type below.  

e.    Others: eimi, go, etc.

II. Reduplicated:

a.    ()stāmi, stand.

b.   (dhé)dhēmi, set, place, do.

c.    ()jēmi, throw, expel.

d.   ()dōmi, give.

e.    (bhí)bheimi, fear.

f.     kkeumi/kuwóm/kékuwa, strengthen.

III. Bisyllabic:

a.    wémāmi, vomit.

NOTE. These verbal types appear mostly in Indo-Iranian and Hittite examples, and could therefore be more properly included in the suffixed (BIVc) type below.  

b.   bhleumi, weaken, (cf. Goth. bliggwan, “whip”).

NOTE. This verb might possibly be more correctly classified as bhleujō, within the verba vocalia, type AIIIu in -u-jo- of the thematic group.

IV. Suffixed:

a.    In -- (<*-neh2): pnāmi, grant, sell (cf. Gk. περνημι, O.Ir. ren(a)id, etc.), qrnāmi, buy (cf. O.Ind. krīnāti, O.Ind. cren(a)im, gr. πρίαμαι, etc).

b.   In -nu-: neumi, rise (up)

c.    With nasal infix: lineqmi (linqō), bhenegmi (bhegō), amneghti (amghō)

NOTE. For these verbs Old Indian shows zero grade root vowel and alternating suffixes.

7.4. The Four Stems

7.4.1. Tense-Stems and Verb Derivation

1. In the earliest reconstructible PIE, secondary verbs existed probably only in the present-tense system, and had no perfect or aorist forms – although presumably they could be conjugated in the imperfect, since it forms part of the imperfective/present system.

NOTE. Even some of the primary verbs were missing perfect and aorist forms, or had forms with unpredictable meanings, and many primary verbs had multiple ways of forming some or all of their aspects. Furthermore, evidence from Old Indian indicates that some secondary verbs in PIE were not conjugated in the subjunctive or optative moods.

Collectively, all of this indicates that in PIH, especially early on, all of the aspects and moods were probably part of the derivational rather than inflectional system. That is, the various tenses, aspects and moods were originally independent lexical formations. Furthermore, a basic constraint in the verbal system might have prevented applying a derived form to an already-derived form (Rix 1986).

That old dynamic situation reconstructed for the PIH conjugation is similar to the system found in Late Indo-European, where old desiderative present stems are generalised as new future stems in a post-LIE period, without the possibility of conjugating it in the subjunctive or optative moods, or even create participles (see below §7.4.2).

2. With verb creation we refer to the way verbs are created from nouns and other verbs by adding suffixes and through reduplication of stems.

3. There are generally two kinds of suffixes: root and derivative; they are so classified because they are primarily added to the roots or to derivatives of them. Most of the PIE suffixes (like -u-, -i-, -n-, -s-, etc.) are root suffixes. The most common derivational suffixes are studied in the following sections.

4. Reduplication is a common resource of many modern languages. It generally serves to indicate intensity or repetition in nouns, and in the Proto-Indo-European verb it helped create present stems (especially intensives), and more frequently it marked the different stems, whether present, aorist or perfect.

5. Examples of the stems found for PIE verbal root leiq-, leave, include:

·Present stem nasal li-n-qe/o- (cf. Gk. limpánō, Lat. linquō, -ere, O.Ir. -léici), and also PII athem. li-n-eq-e/o- (cf. Ved. riákti, Av. irinaxti)

·Aorist stem liq-é/ó- (cf. Ved. rikthās, Gk. élipon, Lat. līquī)

·Perfect stem (-)loiq- (cf. Ved. rireca, Gk. léloipen, Goth. laiƕ, O.Pruss. po-lāikt, O.Lith. liekti)

·Desiderative/Future stem leiq/liq-se/o- (cf. Gk. leípsō).

·Causative-Iterative derivative present stem loiq-éje- (cf. Ved. recayati, Lith. laicaũ, laikýti)

Accent-Ablaut In Inflection

Thematic stems, including subjunctives, had fixed accent on the stem.

In athematic stems, the accent usually alternated in PIE, falling on the endings in the middle-passive and the nonsingular active, but on the preceding syllable in the singular active.

However, s-aorists seem to have had fixed accent on the root, and it appears that there were a few root presents that exhibited a similar pattern; and reduplicated presents (but not perfects) seem to have had a fixed accent on the reduplicating syllable (Ringe 2006).

NOTE. No matter what the accentual pattern was, there was normally a difference in ablaut between the singular active and all other forms of athematic stems. The commoner attested patterns are exemplified in the paradigms shown in §§7.8.1, 7.8.2.

Obviously, the inflection of thematic stems was simpler and easier to learn. In the development of Northwestern dialects, nearly all presents would become thematic.

7.4.2. The Present Stem

Present Root Stem

A pure root stem, with or without thematic vowel, can be used as a present, opposed to the aorist and perfect. Present verbal roots may be athematic and thematic. The athematics were, in Late Indo-European, only the remains of an older system.

Class BIaMonosyllabic Athematic

[LIV types 1a & 1c, Old Indian 2nd class] Monosyllabic athematic root presents ending in consonant or resonant; their inflection is usually made:

·  in the active voice sg., root vowel e and root accent.

·  in the active voice pl. and middle voice, root vowel and accent on the ending.

The most common example is es-mi, to be, which has a singular in es- and plural in s-. There are also other monosyllabic verbs, as chen-mi, to strike, ed-mi, to eat, wek-mi, to will, etc.

NOTE. There was a general tendency within Late Indo-European to use thematic verbs instead of the old athematic ones. “The athematic verbs have been largely replaced by those of the thematic type” Beekes (1995).

BIe.- Other monosyllabic athematic root stems, as ei-mi, go; these follow the same declension.

 

 

ed-, eat

chen-, knok

ei-, go

es-, be

Sg.

1.

edmi

chenmi

eimi

esmi

2.

edsi

chensi

eisi

esiii

3.

estii

chenti

eiti

esti

Pl.

1.

dmes

chmés

imés

smes

2.

dte

ch

i

ste

3.

denti

chenti

jenti

senti

i post-LIE ésti(<*etsti?)<*édti.  ii please note PIE es- + -si = esi.

NOTE. In an old inflection like that of the verbal root es, i.e. sg. esmi, pl. smés, sometimes a semithematic alternative is found. Compare the paradigm of the verb be in Latin, where zero-grade and o vowel forms are found: s-omi (cf. Lat. sum), not es-mi; s-omos (cf. Lat. sumus), not s-me; and s-onti (cf. Lat. sunt), not s-enti. Such inflection, not limited to Latin, has had little success in the Indo-European verbal system, at least in the older IE languages attested. There are, however, many examples of semithematic inflection in non-root verbs, what could mean that an independent semithematic inflection existed in PIE, or, more likely, that old athematic forms were remade and mixed with the newer thematic inflection (Adrados–Bernabé–Mendoza 1995-1998).

The middle voice forms that correspond in terms of formation are treated separately as zero grade root statives in LIV (type 1c); only two cases are certain.

Class BIII – Disyllabic Athematic

Disyllabic athematic root stems that make the present in full/ root vowel; as, wémāmi, vomit, bhleumi, weaken.

The alternative /full root vowel is generally reserved for the aorist.

Classes BIb & BIc – Narten Present

[LIV type 1b, Narten present] Root athematic stems with alternating long/full root vowel and fixed root accent, acrodynamic; as, stēu-ti, 3rd pl. stéw-ti. They are conjugated frequently in the middle voice.

Class AI – bhárati/tudáti

[LIV types 1n & 1o – Old Indian 6th Class] Thematic root stems with e grade and thematic suffix -e/o- before the endings. A common example is bhér-e-ti, ‘bears’. Thematic inflection shows two general old paradigms:

a. Root vowel e and root accent, as déiketi, (s)he/it shows.

b. Root vowel and accent on the theme vowel.

NOTE.  The b-types are called tudáti-presents from a representative example in Ved. tudáti ‘beats’ < *(s)tud-é-ti, a verb that forms  a common LIE nasal infix present (s)tu--d-. There seems to be no certain present reconstructible for common PIE, according to Beekes (1995).

The a-type appears usually in the present, and the b-type in the aorist – pres. déiko vs. aor. dikóm –, although apparently both could appear in both stems in PIE. In fact, when both appear in the present stem, the a-type is usually a durative – meaning an action not finished –, while b-type verbs are terminatives or punctuals – meaning the conclusion of the action. This semantic value is not general, though, often found only in Graeco-Aryan dialects. The a-type present stems correspond to LIV type 1n, while the b-type present stems correspond to LIV type 1o.

NOTE. The newer inflection is, thus (in a singular/plural scheme), that of full/full vocalism for present, / for aorist. The athematic root inflection in full/ appears to be older than the thematic one. The thematic inflection therefore probably overshadowed the athematic one by Late Indo-European, and there are lots of examples of coexisting formations, some of the newer being opposed to the older in meaning.

Present Reduplicated Stem

Depending on its formation, present stems may show either full reduplication, sometimes maintained throughout the conjugation, or simple reduplication, which normally consists of the initial consonant of the root followed by -i-.

Depending on its meaning, reduplication may have a general value (of iteration or intensity), or simply opposed values in individual pairs of basic verb vs. deverbative, helping to distinguish the verb in its different forms.

Simple reduplication is made:

·  With consonant + i,

- in athematic verbs; as, bhi-bher-, carry (from bher-),

- in thematic verbs; as, gi-gnō-sko-, know (from gnō-), etc. si-sdo-, sit down, (from zero-grade of sed-, sit),

- Some intensives have half full, half simple reduplication, as in dei-dik-, show (from deik-).

- There are other forms with -w, -u, as in leu-luk-, shine (from leuk-, light).

- There are also some perfect stems with i.

·  With consonant + e/ē, as dhe-dhē-, de-dō-, etc.

Simple reduplication in e appears mainly in the perfect, while i is characteristic of present stems. Reduplication in e is also often found in intensives in S.LIE.

Class BII – Athematic in -i-/-e-

[LIV types 1g & 1h] Athematic present with -i- or -e- reduplication. Roots with long vowel (as dhē-, stā- or dō-) are rare in present stems, usually reserved for the aorist. The reconstructed PIH paradigm of stā- is given here for comparison.

 

 

dhē-, do

dō-, give

stā-, stand

*steh2-, stand

Sg.

1.

dhédhēmi

()dōmi

()stāmi

*(sí)steh2mi

2.

dhédhēsi

()dōsi

()stāsi

*(sí)steh2si

3.

dhédhēti

()dōti

()stāti

*(sí)steh2ti

Pl.

1.

dhedhamés

(di)damés

(si)stamés més

*(si)sth2més

2.

dhedhaté

(di)daté

(si)staté

*(si)steh2

3.

dhedhanti

(di)danti

(si)stanti

*(si)sth2ṇti

NOTE. 1) Reduplication didn’t affect the different root vowel grades in inflection, and general accent rules were followed; as, bíbherti-bibhmés, sístāmi-sistamés, etc.

2) Most athematic verbs are usually reconstructed with an accent-ablaut paradigm (as in Sanskrit, or the assumed older situation in PIE), but another simple columnar accent could have been possible, as in Greek, probably from a LIE trend to simplify the system, similarly to the simplified nominal accent-ablaut paradigm; it could read post-LIE dhédhames, dhédhate, dhédhanti, or dídames/dames, sístames/stames, etc.

3) Formal reduplication was optional in Late Indo-European, its generalisation being a Graeco-Aryan feature; as, dédōmi/dídōmi vs. dōmi,gígnōskō vs. gnōskō, etc.

4) Reduplication reconstructed in -e- (cf. for dhe-dhē- Skt. dádhāti, O.Lith. desti, O.C.S. deždǫ, Lat. re-ddo?; for dé-dō-, Skt. dádati) is also found in -i- (cf. for dhí-dhē-, Gk. títhēmi; for dí-dō- Gk. dídōmi, Celtib. didonti). The LIV classifies Greek forms in -i- as from an original PIE -e- assimilated to the -i- class, but there is no certainty in that assumption for all cases, given that reduplication appears not to have been obligatory in PIE. A general reduplication in -e- for dhē- seems to be well established for most languages, though. As Fortson (2004) states: in many examples of the type from the daughter languages, the reduplicating syllable has -i- rather than -e-, as in Vedic Skt. --ti ‘he goes’ and Gk. --mi ‘I give’. This pattern probably spread from thematic reduplicated presents like Gk. gígnomai ‘I become’. Or, after Beekes (1995): “It is unclear when e and when I were used. Skt dádāmi, Gr. dídōmi ‘to give’ perhaps suggest that both forms appeared in the same paradigm.”

Class AVII – Thematic in -i-

[LIV type 1i] Thematic present with -i- reduplication is clearly a secondary development from the LIV type 1h. Common examples are gi-gnō, beget, pi-, drink.

Class BIVdIntensives

[LIV type 6a] Stem formations of the action type ‘intensive’, meaning “repeated bringing about of a state of affairs”, have an (almost) complete reduplication of the root (only an occlusive at the end of the root was not repeated); as, wer-w(e)rt-, to turn, dei-dik-(sk)-, to indicate, qér-qr-, to do again and again, from qer-, cut (off), carve.

Full reduplication, normally found in the present stem, repeats the root or at least the group consonant/resonant+vowel+consonant/resonant. gal-gal-, talk, bher-bher-, endure, d-dr-, m-mr-, whisper, murmur, etc.

Full reduplication is also that which repeats a root with vowel+consonant/resonant; as, ul-ul-, howl (cf. Lat. ululāre).

NOTE. Examples include Greek πορφυρω, παμπαινω, γαργαιρω, μορμορω, μερμηριζω, καγχαλαω, μαρμαιρω, δενδιλλω, λαλεω, and, in other IE dialects, Slavic glagoljo, Latin (‘broken’ reduplication with different variants) bombico, bombio, cachinno, cacillo, cracerro, crocito, cucullio, cucurrio, curculio, didintrio, lallo, imbubino, murmillo, palpor, pipito, plipio, pipio, tetrinnio, tetrissito, tintinnio, titio, titubo, etc.

Present Consonant Stem

In -s-

A present-tense thematic suffix -(e)s- is found, for example in kleu-sō, obey, be obedient, g-esō, carry (*h2g-es-, from *h2eg- agō), aug-sō, grow (cf. Gk. aéksō, Lat. augeō).

Thematic -s- also makes desideratives which are the basis of post-LIE futures, v.i.

Extended -s- stems, as -sk- and -st-, are almost all thematic.

NOTE. Thematic suffix -ste/o- has usually an expressive meaning, meaning sounds most of the times; as, bhstō, burst, break (from bhresjō, shatter).

Class AV – In -ske/o-

[LIV type 1p].- Thematic suffix -ske/o- is added to roots in the zero-grade, especially to monosyllabics and disyllabics, and make iterative (or inchoatives); as, pk-sk (from prek-), ask, ask repeatedly, c-sk, walk about (cf. cemjō, come), gnō-skō (from gnō-), know.

It can also be added to reduplicated stems, as -dk-skō (from dek-), -gnō-skō, and to lengthened roots, especially in ī, u, ē, ā, as krē-skō (from ker-).

NOTE. Several verbs reconstructed for PIE with this ending refer to asking or wishing. Sometimes these deverbatives show limited general patterns, creating especially iteratives (with repeated, habitual or background action, i.e. durative sens, cf. Hitt. wal-iški-zzi ‘beats repeatedly, beats several objects’, Gk. pheúgeskon ‘they would habitually flee’), but also inchoatives (indicating beginning or inception of an action or state, cf. Lat. rubē-sc-ere ‘to grow red’), causatives, and even determinatives or terminatives. Apparently, the same -ske/o- can also produce denominal duratives like medhuskō, get drunk (from medhu, mead, intoxicating drink) or wodskō,wash (from wod-, water). (Piotr Gąsiorowski, n.d.)

This lengthening in -sk- seems to have been part of present-only stems in LIE; cf. Lat. flōrescō/flōruī, Gk. κικλησκω/κεκληκα, and so on. Cases like LIE verb pksk, ask, demand (cf. O.H.G. forscōn, Ger. forschen, Lat. poscō>por(c)scō, O.Ind. pcch, Arm. harc’anem, O.Ir. arcu), which appear in zero grade throughout the whole conjugation in different IE dialects, are apparently exceptions of the PIH verbal system; supporting a common formation of zero grade root iterative presents, compare also the form (e)ské/ó- (<h1skó), the verb es- with ‘existential’ sense, as O.Lat. escit, “is”, Gk. ske, “was”, Hom. Gk. éske, Pal. iška, etc.

Supporting the theory that -sk has a newer development than other lengthenings is e.g. the Hittite formation duskiski(ta) (cf. O.Ind. túsyate, ‘silenter’, O.Ir. inna tuai ‘silentia’), which indicates that in Anatolian (hence possibly in Indo-Hittite as well) such an ending – unlike the other endings shown - was still actively in formation.

Classes BIVb & AVI – Nasal Presents

[LIV type 1k – Old Indian class 7] Stems in -n- are said to have a nasal suffix or a nasal infix – a morpheme placed inside another morpheme. They may be athematic or thematic, and the most common forms are -n, and extended -neu-/-nu-, -nā-.

The so-called nasal presents are thematic and thematics with nasal infix -n-, typically active transitives. The distribution of the ablaut grades was the same as in root presents: full grade in the singular active, zero-grade elsewhere.

The infix was inserted into the zero-grade of the root, between its last two sounds (generally a resonant or high vowel followed by a consonant), i.e. in CeRC- roots, the produced the characteristic alternation CR-né-C-/CR-n-C-; as, from jug-, 3rd sg. nasal present 3rd sg.  ju--g-ti, he yokes, 3rd pl. ju-n-g-énti; ku-n-és-mi, kiss.

Other examples include li-ne-q-mi, becoming thematic li-n-q-ō, leave; other thematics include pu-n-g-ō, prik, bhu-n-dh-ō, be aware. Other stems with nasal infix became verba vocalia; as, dhre-n-g-ājō, hold; pla-n-t-ājō, plant.

Classes BIVc & AVI – In -neu-

 [LIV type 1l – Old Indian classes 5, 8] Athematic nasal infix present -néu-/-nu- usually enforcing the weak vocalism of the root, as in st-neu-mi, becoming thematic ster-n(u)-ō, spread, -neu-mi, set into motion, etc.

NOTE. Derivative kneumi is difficult to reconstruct with certainty; often interpreted as with infix -n-, i.e. k-n-eu-, it has been proposed that it is a zero-grade suffixed klu-neu-, cf. Buddh. Skt. śrun; Av. surunaoiti; Shughni çin; O.Ir. cluinethar; Toch. A and B käln. Skt. śRno-/śRnu- < *kluneu-/klunu- would show a loss of u analogous to the loss of i in tRtī́ya- ‘third’ < IE tritijo-.

It produced (often transitive and vaguely causative) athematic verbs that refer to the beginning or termination of an action (the so-called inchoatives), or suggest that something is done once (rather than repeated) (Piotr Gąsiorowski, n.d.).

A rarer variant of this pattern involves -nu-, -ne/o-, formations with stress alternating between the full-vowel root and the inflection.

NOTE. Other forms (possibly derived from inflected -neu- and -nei-) include -nwe/o-, -nje/o-. These formations seem to be very recent in Late Indo-European. In Greek it is frequent the nasal suffix -an-. Others as -nwe/o-, -nje/o-, appear often, too; as Gk. phthínuo, Goth. winnan (from *wenwan); Gk. iaíno, phaínomai (from bhā-) and O.Ind. verbs in -nyati.

Class BIVa – In -nā-

[LIV type 1m – Old Indian class 9] Athematic nasal infix --; as, p--mi, grant, sell, qr--mi, buy, d-nā-mi, to subdue, etc.

In occlusive

Indo-European roots could be lengthened with an occlusive to give a verb stem, either general or present-only. Such stems are usually made adding a dental -t-, -d-, -dh- (as plek-tō, plait, from plek-, weave) or a guttural -k-, -g-, -gh- (as dha-k-jō, do), but only rarely with labials or labiovelars. They are all thematic, and the lengthenings are added to the root.

Present Vowel Stem

Class AIV – Primary je/o-Presents

[LIV types 1q & 1r – Old Indian 4th class] Some roots and derivatives (deverbatives or denominatives) form the thematic verb stems with -je/o-, usually added to stems ending in consonant. These are called primary je/o-presents (not to be confused with “primary” in the meaning “non-past” of verbal endings).

NOTE. According to the LIV, it forms thematic durative verbs, conveying “a subject’s state of being without stressing the entry of the subject into the state of being”; as kapjō, take, seize, msjō, not heed, ignore (from mors-, forget), oqjō, eye (from noun oqos, eye, cf. oqō, see).

In these cases, the root grade is usually ; as, mn-j, from men-, think, bhudh-jō, wake up, from bheudh-; but the full grade is also possible, as in spek-jō, look, lā-jō, from lā-, bark.

NOTE. Fortson (2004): “The type with zero-grade of the root and accented suffix, characteristically used with intransitives, may have been restricted to middle inflection originally, which would explain why in some branches (Indo-Iranian, Armenian) it came to be used to form the passive.”

These verbs may be deverbatives – normally iteratives or causatives – or denominatives. With an iterative-causative action type [LIV type 4b], cf. swopjō, lull to sleep, from swep-, sleep.

They served especially to form verbs from nouns and adjectives, as, from nōm, name, nōmn, name (cf. Gk. onomainō, Got. namnjan), from melit, honey, mitjō, take honey from the honeycomb (as Gk. blíttō), etc.

NOTE. Equivalent stems in thematic -u-e/o- are rarely found in the present, but are often found in the past and perfect stems. Stems in -u- show then an opposed behaviour to those in -i-, which are usually found in present stems, and rarely in past or perfect stems. In present stems, -u- is found in roots or as a suffix, whether thematic or athematic, giving a stem that may normally appear as the general stem of the verb. It is therefore generally either part of the root or a stable lengthening of it (cf. gheu-/ghō-, pleu-/plō-, etc.).

Class AIII – Verba Vocalia

[LIV types 1q & 1r – Old Indian 4th class] The preceding vowel may be an -ā-, -ē-, -i- or -u-, sometimes as part of the root or derivative, sometimes as part of the suffix. Possible suffixes in -je/o- are therefore also the so-called verba vocalia, -je/o-,-je/o-, -íje/o-, and -úje/o-.

Class AIIIa – Factitives

[LIV type 7] Roots or stems in -ā- (<*-eh2-/*-h2-), added to the weak form of a root to produce athematic or thematic stems mixed with -i-, generally indicating “the entry of the subject into a new state of being”; as, am-ā-, love, sēd-ā-, settle (cf. sed-ejō).

NOTE. Athematic presents in -ā- are classified in LIV as “fientive stems”, like mnā-, become furious, from men-, hold a thought.

Some find apparently irregular formations as Lat. amō, “I love”, from an older am-je/o-, mixed with -i-; however, they are sometimes reconstructed as from *amā-, i.e. in -ā without ending (cf. Lat. amas, amat,...), as in Adrados–Bernabé–Mendoza (1995-1998); against it, compare common IE formations as Umb. suboca ‘invoke’, Russ. délaiu, and so on.

Added to thematic adjectival stems it was used to form factitives, verbs meaning ‘to make something have the quality of the adjective’, especially when opposed to statives in -ē- (cf. Hitt. marša-marše-, Lat. clarāre-clarēre, albāre-albēre, nigrāre-nigrēre, liquāre-liquēre); as, from new-o-, new, new-ā-jō, make new,

They may also form statives or duratives. But there are also many deverbatives in -ā- without a special value opposed to the basic verb.

NOTE. Stems in -ā- help create (usually athematic) subjunctives and aorists. -ā- is less commonly used than -ē- to make iterative and stative deverbatives and denominatives.

Class AIIIe – Statives

[LIV type 8] Thematic stems in -ē-, mixed with -i-. Sometimes the -ē- is part of the root, sometimes it is a suffix added or substituting the -e- of the stem.

NOTE. These stem formations are defined in LIV as of the “essive” action type, conveying “a subject’s state of being…without stressing the entry of the subject into the state of being” (see above Class AIV). Its stem ending is reconstructed as *-h1-, as a derivative of *--, upon the fientives with *-eh1-/-h1-.

They may form verbs of state (or statives) if added to and adjectival root thematic in e/o, meaning ‘have the quality of the adjective’, as rudhējō, be red, albhējō, be white, with a stative value, lubhējō, be dear, be pleasing, senējō, be old, etc.

It is also found in combination with -s- in -ē-s-, -ē-ske/o-, yielding intransitive verbs denoting change of state (‘become X’); as, roudhēskō, turn red, senēskō, get old (Piotr Gąsiorowski n.d.).

Class AIIIoCausative-Iteratives

[LIV type 4a] The co-called causative-iterative stems show root in o-grade and accented thematic suffix in -éjo-, conveying the meaning “a cause of bringing about a state of affairs, or the repeated bringing about of a state of affairs”; as, from sed-, sit, sodejō, cause to sit, from men-, think, monejō, remind, advise; wortejō, cause to turn, from wert-, turn, from wes-, dress, wosejō, clothe, put on clothes, (cf. Hitt. waššizzi, Skt. vāsáiati, Ger. wazjan, Alb. vesh), sedejō, be sitting (cf. sed-, sit), bhoudhejō, wake somebody up (cf. bheudhō, awake), ghejō, incite (cf. gujō, reason, discuss), etc. And it is also used to form denominatives, as wosnejō, buy, sell, from wesnom, sale.

It formed non-causatives, too; as, from leuk-, light, loukéjō, shine (cf. Hitt. lukiizzi, Skt. rocáyati, Av. raočayeiti, O.Lat. lūmina lūcent).

NOTE. It is sometimes difficult to know if the original form was -éje/o- or -je/o-, because the former is apparently attested only in Anatolian, Indo-Iranian, Greek and Armenian (cf. Arm. Gen. siroy, “love”, sirem, “I love”<*keire--); Greek loses the -j- and follows (as Latin) the rule ‘uocālis ante uocālem corripitur’, what helps metrics. However, Greek had probably a present with long ē (as in non-liquid future and perfect). Mycenaean doesn’t help clarify uncertain reconstructions; moreover, it is often accepted that some forms in O.Ind. -ayati are isolated. See Appendix II – Guide to the Reader for dubious reconstructions.

Desideratives and The Future Stem

[LIV type 5] Following the LIV, the desiderative action type conveys “the subject’s desire or intent to bring about a state of affairs”. These stems are built with a thematic -s- ending. cf. wéid-se/o-, ‘want to see, go to see’, hence‘visit’, as Lat. vīsere, Goth. gaweisōn, O.S. O.H.G. wīsōn, a deverbative from root weid-, from which the general present stem is wid-je/o-, see.

NOTE. Aorist stems in -s- are usually athematic. Sometimes the -s- marked the subjunctive. Because of its common use in verbal inflection, deverbatives with a lengthening in -s- aren’t generally opposed in meaning to their basic stems, and there was no general common meaning reserved for the extended stem in -s-. Compare Lat. pressī <* pres-sai vs. Lat. premō; Lat. tremō vs. a Gk. τρεω<*tre-, O.Ind. trásate ‘he is frightened’.

Some of their descendants function as futures, hence the assumption that future formations in IEDs come from LIE desideratives/causatives. Present stems, usually formed with extensions in -s- (and its variants), became with time a regular part of the verbal conjugation in some dialects, whilst disappearing in others.

NOTE. It is assumed then that PIE did not have a future stem. That might seem strange, but it is possible to express future tense without having a special formation (‘I do it tomorrow’). Gothic, for example, is an IE language that didn’t have an especial future formation. Nevertheless, the development of the earliest languages attested show that within the post-LIE times a future stem must have been developed.

[LIV type 5a] Desiderative/causative stems were usually made in IEDs with root vowel e,  i.e. in full-grade, with a suffix -s-<*-(H)s-:

1. Thematic -s(j)e/o-; as, do-sjé-ti, ‘he intends/wants to give’, later ‘will give’ or ‘about to give’ (cf. Skt. dā-y-mi, Lith. dúosiant-).

NOTE. A common origin of the future in -s- is found in Sanskrit, Balto-Slavic, Italic (Sabellian), and in Celtic futures in -sje/o- (cf. O.Russ. byšęš<*bhuH-sjont- ‘about to be’, Gaul. pissíiumi ‘I will see’), and Doric Greek in -sēje/o-, -sje/o-. Cf. also Hom. Gk. kaléō <*kal-e- ‘I will call’, Classical Greek and Archaic Latin in -se/o- (cf. O.Lat. faxō<*dhak- ‘I will make’, O.Lat. peccas-, from peccāre, etc. and Gk. dk-s-o-mai ‘I will bite’ to active present dáknō ‘I bite’). Cf. from derk-, see, Skt. drakyáti ‘he shall see’, and Gk. dérksomai ‘I shall see’. Some more dialectal extensions are found appearing before the -s- endings; as, -i-s- in Indo-Iranian and Latin, -e-s- in Greek and Osco-Umbrian.

For the future stem coming from sigmatic aorist stem, Adrados–Bernabé–Mendoza (1995-1998): “Homeric Greek aorists dúseto, bseto, are exactly parallels to future dúsomai, bsomai, remains of the same sigmatic thematic stem, and not remade forms as Leumann (1952-53) and Prince (1970) proposed.”

2. An athematic future in -s- is found in Italic (cf. Umbr. fu-s-t ‘he will be’) and Baltic (Lith. bùs ‘he will be’).

3. [LIV type 5b] A reduplicated desiderative with i-reduplication and a suffix *-(h1)se-, found in Indo-Iranian and Celtic; as, wi-w-sō, overpower, win, from wen-, overpower, win; from chen-, slay, chi-chnā-se-ti<*gwi-gwn̥-h1se, wants to slay, will slay (cf. Ved. Skt. jíghāsati ‘wants to slay’, O.Ir. fut. (-)géna < Cel. *gwi-gwnā-se-ti ‘will slay’); di-dk-sō, want to see.

It seems that future stems originated already within a disintegrating post-LIE community, which tended to integrate the known -s- desiderative present stem formation paradigm into the conjugation system, so that it became possible to create systematic futures of all verbs.

Imperfect and Injunctive

The present stem was used to form the present tense and the imperfect, which –as already said – is usually thought to have signified durative or repeated action in the past time (was going, used to go). Formally it was usually identical to the present stem, except that secondary endings were used instead of primary. 1st sg. -m is the same in both thematic and athematic imperfects.

NOTE. Fortson (2004) continues: “The original type is best preserved in Anatolian, Indo-Iranian, and Greek: for example, Hitt. (preterite) daškinun ‘I (repeatedly) took’, Vedic Skt. ábharam ‘I was carrying’, Av. barəmz ‘I was carrymg’, and Gk. épheron ‘I was carrying’ (…) Outside of these branches, the IE imperfect has either been completely lost, or merged with the aorist. In those branches where the imperfect was lost, a new imperfect conjugation was often innovated (as in Italic and Slavic), sometimes of obscure origin (as in Celtic).”

Indo-Iranian, Greek, Armenian and Phrygian attest a prefix called the augment that was added to past-tense forms. It is reconstructible as e-; as, imperfect é-bher-e-t ‘he was carrying’ (cf. Ved. Skt. ábharat, Gk. éphere, Arm eber), or aorist e-dhē- ‘placed’ (cf. Phrygian edaes ‘he placed’).

NOTE. The great success of that particular augment (similar to other additions, like Lat. per- or Gmc. ga-) happened apparently later in those proto-languages. Vedic Sanskrit clearly shows that augment was not obligatory, and for Proto-Greek, cf. Mycenaean do-ke/a-pe-do-ke, Myc. qi-ri-ja-to, Hom. Gk. πριατο, etc. It is often shown in most PIE grammars because (Brugmannian) tradition in IE studies has made augment seem obligatory for PIE.

According to Meier-Brügger (2003): “The PIE augment *(h1 was quite probably an adverb with the meaning ‘at that time’ and could be employed facultatively where indicative forms of present and aorist stems were combined with secondary endings to produce a clear past tense (…) The establishment of the augment as a norm in the indicative aorist, indicative imperfect, and indicative pluperfect took place in a post-Proto-Indo-European phase. Other IE languages such as Latin or Germanic developed their own suffixal means of indicating past tense forms.”

The augment is in fact related to the so-called ‘injunctive mood’, defined as augmentless past-tense forms that appear in Indo-Iranian and Ancient Greek, of obscure function, much discussed by scholars. “Their precise function or functions are still not fully clear. In Homer, injunctives are interchangeable with past tenses but sometimes have gnomic force (that is, are used to express general truths). In Indo-Iranian, injunctives can indicate intent, futurity, and some quasi-modal meanings, and were also used in commands, especially prohibitions” (Fortson 2004).

7.4.3. The Aorist Stem

Aorist Root Stem

[LIV type 2a] Monosyllabic athematic root aorists are formed by adding the secondary endings directly to the full grade of the root in the active singular, and to the zero-grade of the root elsewhere. They are usually opposed to presents:

·  In -neu-; as, pres. kneumi, aor. 3rd p. sg. kleum, 3rd p. pl.  klwent, hear, or pres. qneumi vs. aor. qerm, make, do, etc.

·  Reduplicated; as, pres. -stā-mi, I stand, aor. 1st sg. stā-m, I stood, 3rd pl. sta-nt, they stood; pres. dhé-dhē-mi, I do, I put, aor. dhē-m, I did, pres. (m)-plē-mi, I fill, aor. plē-m, I filled

·  In -ske/o-, -je/o-; as, pres. csk, I walk about, aor. 3rd sg. cem-t, he walked about, 3rd pl. cm-ent, they walked about.

·  Thematic presents; as, pres. ghewō, I pour, aor. gheum, I poured.

Thematic aorist root stems are usually made in / root vowel grade, secondary endings, and sometimes reduplication; as, pres. deik-ō, aor. dik-óm, pres. linq-ō, I leave, aor. liq-óm, I left, pres. wid-, aor. wid-óm; cf. also from leudh- go/come, ludhóm, I went, I came, most commonly used as aorist of cemjō (<cjō), I come.

NOTE. As already seen, these stems could form aorists and presents: The liqé/ó- form (i.e. zero grade and accent on thematic vowel) is usually reserved for the aorist stem, while the leiqe/o- form (i.e. full grade) is rarely found in the aorist – but, when it is found, the present has to be differentiated from it. This is made (1) with vowel opposition, i.e., full grade, o-grade or zero grade, (2) thematic vowel, or (3) with secondary phonetic differentiations (as accent shift).

Aorist Reduplicated Stem

[LIV type 2c] Aorist reduplicated stems are usually thematic, with a general vowel e (opposed to the i of the present), zero-grade root vowel (general in aorists); as, chenmi/che-ch-om, murder, kill; weqmi/we-uq-om, say, speak;

NOTE. Fortson (2004): “(…) reduplicated aorists typically have causative meaning, such as Ved. Skt. á--par-as ‘you made cross over’, Gk. -da-e ‘he taught’ (<’caused to know’), and Toch. A śa-śärs ‘he made known’.” Cf. also Lat. momorit, totondit, spopondit, etc., or O.Ind. atitaram, ajijanam, etc.

In roots which begin with vowel, reduplication is of the type vowel+consonant.

Some roots which begin with vowel form also reduplicated aorists; as ag-ag-om (as Gk. ηγαγον, where η<ā<*é+a – Wackernagel, hence *é-agagom).

Aorist Consonant Stem

[LIV type 2b] The most common consonant stem is the sigmatic aorist, formed with the suffix -s-, generally athematic.

The -s- is added usually to the root, whether monosyllabic or disyllabic, in consonant or vowel, opposed to the present. Such root aorists usually show lengthened e-grade in the active voice, and zero grade in the rest; as, pres. (m)plēmi, I fill, aor. plēsm, I filled, 3rd pl. plest, they filled; qērs, I made, 3rd pl. qst, they made, from qer-, make; dēik-s-m, I indicated,  wēgh-s-m, I carried, I conveyed, etc.

NOTE. Lengthened vocalism in sigmatic aorists was probably an innovation in Late Indo-European. For lengthened grade, cf. maybe Latin forms like dīxī (<*dēik-s-), uēxī from uehō (cf. O.Ind. ávāk-am from váhāmi ‘drive’), rēxī from regō, etc., or Toch. B preksa, A prakäs (<*prēk-s-ā), according to Lindeman (1968). Without lengthening (i.e. full grade) they are found in Greek and in the s-aorist middle in Indo-Iranian, cf. Gk. élekse<*é-leg-s- ‘he said’. For Beekes (1995), the lengthened grade “perhaps it has disappeared through regular sound developments (*dēik- deik-, Osthoff’s Law).”

It could also be added to a vowel ā, ē, ō, with the same stem as the present, or to the noun from which the verb is derived; as, pres. alkē, aor. alkēsom, grow.

The general system of thematic present vs. sigmatic aorist stems may be so depicted: -ēje/o- vs. -ēs-; -āje/o- vs. -ās-; -je/o- vs. -is-; -je/o- vs. -ās-; -je/o- vs. ēs-; and -e/o- vs. -ās-.

Monosyllabic or disyllabic sigmatic aorist root stems in i, u, ā, ē, ō, have a fixed vowel grade; as, aor. pewism, pewist, purify.

NOTE. Aorist stem formation in -i-, -ē-, -ā- is still less common. Other common formations in -s- include the following: In -is- (Latin and Indo-Aryan), -es- (Greek), as genis- from gen-, beget, wersis- from wers-, rain; also, cf. Lat. amauis (amāuistī, and amāuerām<*-wisām), etc. In --, attested in Latin, Tocharian and Armenian. Also attested are aorists in --, thematic -sje/o-, etc.

Stems in -t- function usually as aorists, opposed to present stems, especially in Italic, Celtic and Germanic.

NOTE. While the use of -t for persons in the verbal conjugation is certainly old, the use of an extension in -t- to form verbal stems seems to be more recent, and mainly a North-West IE development.

Stems in -k- are rare, but there are examples of them in all forms of the verb, including aorists.

Aorist Vowel Stem

Aorists in ā, ē, are very common, either as root stems with athematic inflection, or mixed with other endings, e.g. -u-.

NOTE. As already said, stems extended in -u- are rarely found in present stems, but are frequent in past stems; the opposite is true for -i-.

When opposed to a present, stems extended in -ā, -ē, are often aorists. Possible oppositions present stem vowel vs. aorist stem vowel include:

·  Present thematic in -i- vs. aorist athematic in -ē, -ā; as, m vs. mm, consider, alkē vs. alkm, protect.

·  Present thematic in e/o vs. aorist athematic in -ē, -ā; as, le-legēm, collect.

The use of stems in -u- is usually related to the past, and sometimes to the perfect. Such endings may appear as -we/o-, often -āwe/o-, -ēwe/o-; as, plēu-, from plē-, sēu-, from -, gnōu-, from gnō-.

Endings -i-/-ī- are scarcely used for aorists, but they appear in some stems used both for present and aorist stems; as, awisdhijō vs. awisdhijóm, hear, Lat. audĭo, audĭui.

7.4.4. The Perfect Stem

[LIV type 3a] The perfect stem has or lengthened root vowel and special perfect endings, sg. -a, -tha, -e, pl. -, -, -()r, which are only used in the perfect indicative.

In Gk. and Ind.-Ira., the stem was often reduplicated, generally with vowel e; in Latin and Germanic, reduplication is often absent.

NOTE. Historically the perfect was possibly a different stative verb, a deverbative from the root with certain formation rules, which eventually entered the verbal conjugation, meaning the state derived from the action of the present stem. PIE perfect did not have a tense or voice value.

Root vowel is usually /, i.e. o-grade in the singular and zero-grade in the plural; for a contrast pres. 1st sg./ perf. 1st sg. / perf. 3rd pl., cf. gígnō / -gon-a / ge-g-, know; bhindh-ō / bhondh-a / bhdh-, bind; bheudhō / bhoudh-a / bhudh-, wake up.

NOTE. 1) for different formations, cf. kan-ō / ()kan-a / k-, sing, cf. O.Ir. cechan, cechan, cechuin (and cechain), cechnammar, cechn(u)id, cechnatar.; d-ō-mi / de-d-ai, give, cf. O.Ind. dadé, Lat. dedī. 2) For examples of root vowel ā, cf. Lat. scābī, or Gk. τεθηλα, and for examples with root vowel a, cf. Umb. procanurent (with ablaut in Lat. procinuerint) – this example has lost reduplication as Italic dialects usually do after a preposed preposition (cf. Lat. compulī, detinuī), although this may not be the case  (cf. Lat. concinuī).

2) There are also perfects with lengthened root vowel; as, from Latin sedē-, perf. sēd-a, sit; ed-ō, perf. ēd-a, eat; cem-jō, perf. cēm-a, come; ag-ō, perf. āg-a, act; from Germanic, sleb-ō, perf. séslēb-a, sleep; etc.

Reduplication is made in e, and sometimes in i or u.

NOTE. Apparently, in Indo-Iranian and Greek dialects reduplication was obligatory, whereas in North-West Indo-European it wasn’t. For an older nonobligatory reduplication, there is common PIE perfect woistha (<*woid-th2e), know, from weid-, see (hence the stative meaning ‘state derived from having seen’?), cf. O.Ind. véttha, Gk. (w)oīstha, Goth. waist. Cf. also afield Gk. εγνωκα, Lat. sēuī (which seems old, even with Goth. saiso), Lat. sedī, from sedeō and sīdo, which do not let us reconstruct whether the original form is sesdāi or sēdāi.

7.5. Mood Stems

7.5.1. Indicative

The indicative expresses the real action, and it is the default mood; the other ones were specialised in opposition to it. It appears in the three verbal stems. The following table depicts the minimal verbal stem system attested for IEDs, according to finds from Vedic Sanskrit and Ancient Greek (Clackson 2007):

 

Present

Aorist

Perfect

Future? ii

Indicative

past & non-past active & middle

past                    active & middle

[no voice/tense opposition]

non-past    active & middle

Subjunctive

active & middle

active & middle

[no voice opp.]i

-

Optative

active & middle

active & middle

[no voice opp.]i

-

Imperative

active & middle

active & middle

-

-

Infinitive

active & middle?

active & middle?

-

-

Participle

active & middle

active & middle

[no voice opp.]

-

i The moods in the perfect were probably developed late in the history of the proto-language (see §7.4.1). Adrados–Bernabé–Mendoza (1995-1998), Fritz (Meier-Brügger, 2003), Fortson (2004), Clackson (2007), among others attribute their development to a late innovation, probably a Graeco-Aryan one. For a reconstruction of a PIE moods in the perfect, see Beekes (1995), Ringe (2006). Given that Northwestern dialects tended to merge later perfect, aorist and imperfect into a common preterite, it is probably safe to assume that in any case mood distinctions for the perfect were not much used in NWIE.

ii With the future we refer to the post-LIE stem incorporated into the conjugation of a certain verb. Those desideratives with present stems in thematic -s- reconstructed for LIE (of which the LIV cites more than 100 examples, with almost 40 of them certain) could have in turn their own present, aorist, perfect, and even future stems.

The general mood system might be so depicted:

 

Indicative

Imperative

Optative

Subjunctive

Function

action described as fact

commands

wishes, hopes

action described as completely theoretical

Charact.

default mood

not conjugated in the first person

special personal endings

ablauting athematic  -ī/jē-, thematic -oi- affixed to stem

secondary endings

thematic -e/o- suffix affixed to stem

primary endings

 

7.5.2. Imperative

The imperative, used to express direct commands, had probably in PIH the same basic stem of the indicative, and was used without ending, in a simple expressive-impressive function, of exclamation or order. Imperatives are the equivalent in verbal inflection to the vocative in nominal declension.

NOTE. In Late Indo-European a new system was derived from this older scheme, a more complex imperative system, featuring person, tense and even voice.

The general athematic 2nd sg. imperative ending was -; as in ei! go! from eimi; or es! be! from esmi. An ending -dhí seems to have been common in LIE too; as, i-dhí! go!, s-dhí! be!

The thematic 2nd singular imperative was the bare thematic stem; bhere! carry!, age! do! act! The 2nd pl. ends in -te, as bhérete! carry!, agete! do! act!

The thematic and athematic 3rd sg. and 3rd pl. have a special ending -tōd.

NOTE. Endings in *-u, i.e. 3rd sg. *-tu, 3rd pl. *-ntu, are also reconstructed (see Beekes 1995) from forms like Hitt. paiddu ‘let him go’, or Skt. é-tu, ‘go’; the inclusion of that ending within the verbal system is, however, difficult. A common IE ending -tōd (cf. Skt. -tt, Gk. -, O.Lat. -tōd, Celtib. -tuz, Goth. -dau), on the other hand, may obviously be explained as the introduction into the verbal conjugation of a secondary ablative form of the neuter pronoun tod, this, a logical addition to an imperative formation, with the sense of ‘here’, hence ‘now’, just as the addition of -i, ‘here and now’ to oppose new endings to the older desinences (Adrados–Bernabé–Mendoza 1995-1998, Fortson 2004). This formation was further specialised in some dialects as future imperatives.

In root athematic verbs, plural forms show vowel and accent on the ending; as, s-entōd! be they!

For Late Indo-European, only the person distinctions of the active voice are reconstructed with certainty. Common middle forms include the bare stems plus middle desinences; as, 2nd sg. -s(w)e/o (cf. Skt. -sva, Gk. lúou<*lúe-so, Lat. sequere<*seque-se, Ira. *-swe/o), 2nd pl. -dhwe, cf. Gk. lúes-the, O.Ind. bháva-dhvam. Beekes (1995), Sihler (1995).

 

Athem.

Them.

Middle

Sg.

2.

-, (-dhí)

-e

-so

3.

-tōd

-etōd

(-to)

Pl.

2.

-te

-ete

-dhwe

3.

-tōd

-ontōd

(-nto)

NOTE. Forms for the 3rd person are uncertain, although a common sg. -tōd (from the active voice) is reconstructed; cf. Skt. -tāt, Gk. -sthō (sth- from plural and -ō from tōd), Lat. -tōd. Middle secondary endings 3rd sg. -to, 3rd pl. -nto, are tentatively reconstructed by Beekes (1995) as imperative marks with basis on the Sanskrit (-tām, -ntām) and Hittite (-taru, -antaru) endings.

7.5.3. Subjunctive

1. The subjunctive is normally formed by the addition of the thematic vowel to the verb stem (be it athematic or already thematic), followed apparently by primary endings (although in Indo-Iranian both primary and secondary endings were used).

The subjunctives made from thematic verbs end therefore usually in so-called ‘doubled’ thematic vowels, i.e. -ē, -ō, and -ā, always opposed to the indicative. These are sometimes called athematic subjunctives.

NOTE. The reader should take on account that the classification of forms in -ā, -ē, -ō, as “athematic” is purely conventional; so, for example, Dahl considers them as two thematic stem allomorphs, terming them long-vowel-subjunctive [=athematic] against short-vowel-subjunctive [=thematic]. See <http://folk.uio.no/eysteind/PaperICHL.pdf>.

2. The subjunctive always has full grade in the root, and is usually made following these rules:

a.  Indicative athematic vs. subjunctive thematic; as, ind. esmi, I am, senti, they are, subj. esō, (if) I be.

b.  Indicative thematic vs. subjunctive with lengthened thematic vowel; as, ind. bhéresi, you carry, Sub. bhérēsi, you may carry, (if) you carried.

NOTE. Following Meier-Brügger, “[t]he subjunctive suffix is PIE *-e-, In the case of athematic verbal stems, the rule is [where K=Consonant] -K+- (indicative stem), -K+e- (subjunctive stem); correspondingly, that of thematic verbs is -e+ - (indicative stem), -e+e- (subjunctive stem).”

3. Subjunctives could also be formed in the same way from root and s-aorists, where likewise the full grade of the aorist stem was used (Fortson 2004).

NOTE. As indicated by first-person subjunctives like Ved. kr̥avā ‘I will do’, Old Avestan yaojā ‘I will yoke’, Gk. phérō ‘let me carry’, and Lat. erō ‘I will be’, the 1st singular ended in -ō rather than -mi (Fortson 2004).

7.5.4. Optative

The optative mood is a volitive mood that signals wishing or hoping, as in English I wish I might, I hope it may, I wish you could, etc. It is made with the following suffix, and secondary endings.

1) In the athematic flexion, a general alternating suffix -jē-/-ī-  with full-grade in the singular and 3rd pl. and zero-grade elsewhere; as, s-j-m, may I be, s-ī-, may we be, es-ī-nt, may they be.

NOTE. “The stress was on the ending in the 1st and 2nd pl. forms of the mobile paradigms, and evidently also in the sg. forms of the middle voice, but not in the 3rd pl. forms, where a number of indications point to original root stress”, as Lat. velint, Goth. wileina, and O.C.S. velętъ. But, Vedic -ur appears “in all those athematic forms where the stress is either on the root or on a preceding syllable”. See <https://openaccess.leidenuniv.nl/handle/1887/2878>, Kortlandt (1992).

2)  When the stress is fixed, it is -oi- in the thematic flexion, and -ī- in the athematic (e.g. Narten presents); as, bher-oi-t, may he carry.

NOTE. This is probably the thematic -o- plus the zero-grade Optative suffix -i- (<*i-h1-), i.e. originally *-o-ih1-, or maybe *-o-jh1-, see Hoffmann (1976). Optative endings might yield a reconstruction of vocalic resonants in PII, PGk from -o-j, -o-jt.

3)  In the 1st person middle the thematic ending is found (cf. Skt. bruv-ī); as, s-īj-á, bhér-oj-a.

Athematic stems have usually root vowel in zero-grade, while thematic stems show no ablaut.

NOTE. For athematic optatives form the present with zero-grade; cf. Lat. siēm, duim, Gk. ισταιην, διδοιην, τιθειην, O.Ind. syaam (asmi), dvisyām (dvesmi), iyām (emi), juhuyām (juhkomi), sunuykām (sunomi), rundhyām (runadhmi), kuryām (karomi), krīnīyām (krīnāmi), etc. Exceptions are Lat. uelim (not uulim), Goth. (concave) wiljau, wileis, etc.

7.7. Noun and Adjective Forms

7.7.1. Infinitives

1. The infinitives are indeclinable nouns with non-personal verbal functions, which can be as many as inflection, voice, aspect and even time.

2. The oldest infinitives are the verbal nouns, casual forms inflected as nouns, sometimes included in the verbal inflection. A verbal noun is a declinable substantive, derived from the root of a verb.

NOTE. Infinitives are, thus, old nouns reinterpreted as forming part of the verbal conjugation, probably within the Late Indo-European period. As Meier-Brügger (2003) notes, “The development of means of differentiation of voice, aspect, and tempus in the infinitive formations is post-Proto-Indo-European.”

The difference in syntax is important: the verbal noun is constructed as a substantive, thus e.g. with the object in the genitive; as, wīrosjo chentis, the murder of a man. Such a formation is opposed to an infinitive with an accusative; as, wīróm chentum, to murder a man.

3. Verbal nouns were, thus, the normal way to express the idea of a modern infinitive in PIE. They were formed with the verbal stem and usually a nominal suffixes -ti-, -tu-; as, statis (<*sth2-ti-), standing, placing, from stā- (<*steh2-) stand; cem-tus, coming, from cem-, come.

NOTE. Cf. Skt sthíti- ‘stay, sojourn’, Grk stásis ‘place, setting, erection [of a statue]’, Lat statim ‘firmly, steadfastly’, Eng. stead. Mallory–Adams (2007). Some IE dialects chose later between limited noun-cases of those verbal nouns for the infinitive formation, generally Acc., Loc., Abl.; compare Lat. *-os (sibilant neuter), Gmc. *-on-om (thematic neuter),  etc.

4. In Late Indo-European, a common infinitive suffix -tu- (and more limited -ti-) seems to have been usually added to the accented strong verbal root, conveying the same meaning as the English infinitive; as, stātum, to stay, opposed to the weak, unaccented form in participle statós, placed.

NOTE. For generalised IE infinitive formation in -tu [generally -tu-m, i.e. the accusative of the abstract noun suffix -tu-, often called supine (solely used with verbs of motion to indicate purpose)], cf. Skt. -tus, -tum (acc.), Gk. -tós (<*-tew-os), Av. -tos (gen.), -tave, -tavai (dat.), -tum, Lat. (active & passive supine) -tum (acc.) - (dat.-loc.) -tui (dat.), Prus. -twei (dat.) -tun, -ton (acc.), O.Sla. -tŭ (supine), Lith. -tų, etc.; for -ti-, cf. Ved. -taye (dat), BSl., Cel. -ti (loc.), Lith. -tie (dat.), etc.; also, in -m-en-, cf. Skt. -mane, O.Gk. -men(ai), etc.  

Also, a common ending -dhjāi added to the verbal stem formed common middle infinitives.

NOTE. The reconstructed -dhjāi (Haudry), is the basic form behind Ved. -dhyai, Gk. Middle -σθαι, Umb. -fi, Toch. -tsi, as well as Latin gerunds and the Germanic *-dhjōi (Rix 1979), all related to an original middle infinitive (Beekes 1995, Sihler 1995), although appearing in both active and passive formations (Fortson 2004). Other forms include -u-, -er/n-, -(e)s-, extended -s-, -u-, -m-, also Gmc. -no- (as Goth. ita-n<*edo-no-), Arm. -lo-, etc.

7.7.2. Participles

1. The participles are adjectives which have been assimilated to the verbal system, expressing tense and voice; like other adjectives, they have nominal inflection.

NOTE. The reconstructed Proto-Indo-European shows an intense reliance on participles, and thus a certain number of participles played a very important role in the early language.

2. Those in -nt-, fem. -nt-ja/ī, are the older ones, and form participles of active voice to present or aorist stems.

NOTE. In Anatolian, this participle is semantically equivalent to verbal adjectives in --.

In athematics it seems that an ablauting suffix -e/ont with full and zero grade coexisted in the declension of present participles; cf. s-ent-(also s-ont-)/s-t-, who exist, being, weq-ont-/uq-t-, who speaks, speaking, dhe-dhē-nt-/dhe-dha-nt-, placing, jung-ent-/jung-t-, joining, d-ent-, eating (from which dentis, tooth), j-ent-, going, chn-ent-, killing, ag-ent-, driving, guiding, etc.

NOTE. For s-t- instead of sent-, cf. ap-st- (for apo-we-s-ent-is) in Lat. (ab)sent-, Myc. pl. (a-p)e-a-sa, i.e. ap-ehassai (with -assa-<*-t-ih2-). A. Morpurgo Davies (1978, reviewed in Meier-Brügger 2003) considered that “[a]s far as we know, there is no reason to attribute *h1s-ent- to Proto-Greek.”

In thematics, a form -o-nt- (i.e. -nt added to the thematic vowel) is generalised as, bher-ont-, who carries, carrying.

NOTE. The suffix -o-nt- shows no generalised ablaut full-grade/zero-grade paradigm in IEDs. It is safe to assume no accent-ablaut change for North-West IE, and probably also for LIE, as “[i]t remains to be seen whether the thematic forms were originally declined as *-ont-/*-nt- (as in Vedic), and were only secondarily reinterpreted as *-o-nt-”, as some have posited; Meier-Brügger, 2003 (reviewing Rix 1976, Szemerényi 1990).

Also, some questions about the participles are not easily reconciled: in Latin, they are formed with e ending for stems in -i-; in Greek, they are formed in o and are consonantal stems. Greek, on the other hand, still shows remains of the thematic vowel in participles of verba vocalia -ājont-, -ējont-, etc. Latin doesn’t.

Aorist active participles were formed similarly to present participles, as the root aorist participle stā-nt-, having stood (cf. Ved. sthānt-, Gk. stant-), s-aorist dhech-s-t- [‘dhek-sn̥t], having burnt, dik-s-t-, having indicated,

3. The perfect active participle has an ablauting suffix -wos-/-us-, fem. -us-ja/ī; as, weid-wós-, wid-us-ja, knowing, ‘who is in a state of having seen’, from weid-, see; bher-wós-, ‘who is in a state of having carried’. Common is the reduplicated Perfect stem; as, qe-q-wós-, making, from qer-.

For the declension of these participles in -nt- and -wos-, see above §5.2.

4. The middle participles have a common suffix -mo- for athematic, -o-mo in thematics; as, bhéro-mos, carrying (oneself, for oneself), álo-mos, who feeds himself, nurtured, from alō, raise, feed (cf. Lat. alumnus), dhē-mā, the one who gives suck, from dhē-i-, suck (milk), suckle (as Av. daēnu-, Lat. femina, ‘woman’).

NOTE. On the *-mXno- question, where X is a vowel or laryngeal or even laryngeal+vowel, while Melchert (1983) or Szemerényi (1990) support an original -mn-o-, a competing hypothesis is Fritz’s one with an original *-mh1eno-, into variants *-mh1no- and then -mno-, in which “the laryngeal disappears when the suffix is added to a root or stem with a non-syllabic final position preceding the full vowel e. The non-laryngeal full grade form *-meno- would then have the newly constructed zero grade form *-mno-” (Meier-Brügger 2003). The differentiation of the perfect *-mh1n-ó- vs. the present *´-o-mh1no- in the various IE languages may be traced back to the athematic/thematic dichotomy (Rix 1976). For an explanation on the auxiliary vowel in Adrados–Bernabé–Mendoza (1995-1998), see §2.3.

5. In addition to participles, PIE had verbal adjectives in -- and --, added usually to the zero-grade of a verbal stem that indicated completed action, and were semantically like past participles in English. If the verb they were formed from was transitive (like eat), the verbal adjective was passive and past in tense (eaten), but if the verb was intransitive (like go), the verbal adjective was simply past in tense (gone) Fortson (2004). Examples include ch-tós, slain, from transitive chenmi, murder, cf. Skt. hatá-, Gk. -phatós, sjū-tós, sewn, from sisō, t-tós, stretched, klutós, heard; c-tós, (having) come, from intransitive cemjō, come.

a. General --, added usually to zero-grade roots; as, altós, grown, dhatós, placed, kaptós, taken, liqtós, left, etc. Exceptions include e.g. gnōtós, having been understood.

b. Old (not generalised) -- and its variants; as, plēnós, ‘(having been) filled up’, full, bhidhnós, ‘having been split’, parted, bitten; wgnós, worked.

NOTE. For pnós, from pel-, fill, an adjective which was not part of the verbal paradigm, cf. Skt. pūrá-, Lat. plēnus(vs. past participle –plētus ‘filled’),  Goth. fulls (double -ll- < *-ln-), O.Ir. lán, Lith. pìlnas. Also, the common PIE verb is found from this root, plnāmi, fill, cf. O.Ind. pṇtiGoth. fullnan, Ger. füllen, O.Ir. lín(a)im, Arm. lnum, and root Gk. píplēmi.

Verbal adjectives in --, --, functioned as past participles in individual languages; as, present passive participle in Balto-Slavic -mo-, cf. O.C.S. nĕsomŭ, Lith. šamas ‘being carried’, perhaps Anatolian, cf. Luv. šammi- ‘combed’. For its old use, cf. prāmós, foremost, first, from per-, v.s. §5.5.2; however, Latin prīmus is usually reconstructed as from prei-isamós (cf. Paelignian prsmū), but possibly superlative pw-isamós, from the same root as common PIE prāmos, prāwos, first, is the solution (see Szemerényi 1970, Adrados–Bernabé–Mendoza 1995-1998).

7.7.3. Gerundives and Absolutives

1. Verbal adjectives are not assimilated to the verbal system of tense and voice. Those which indicate need or possibility are called gerundives.

NOTE. Verbal adjectives and adjectives (as verbal nouns and nouns) cannot be easily differentiated.

2. Whereas the same participle suffixes are found, i.e. --, --, --, there are two forms especially identified with gerundives in IEDs:

a. -- and -li- are found in Latin, Balto-Slavic, Tocharian and Armenian; as, bherelós, unbearable, ágilis, agile, etc.

NOTE. For suffix -lo- as a participle suffix, cf. Russ. videlŭ, Lat. credulus, bibulus, tremulus, etc.

b. -- (a common lengthening to differentiate adjectives) is sometimes a gerundive of obligation, as well as forms in -tu-, -ti-, -ndho-, etc.; as, dhsjós, that has to be dared; gnōtinós, that has to be known; gnskendhos, that has to be born, awisdhíjendhos, that has to be heard; and so on.

NOTE. Some forms in -ndhos seem to retain a so-called fossil proto-gerundive (Meiser 1998), from an archaic ending *-dnós, whose meaning lack the passive obligation common to the gerundive; so e.g. *mlāje-dnós>O. Lat. *blandos, *rotodnós (Lat. rotundus), round, or *seqodnós (Lat. secundus). Outside Latin it is possibly found e.g. in Gk. tēked ‘consumption’ or *phagedn (cf. Gk. phagédaina ‘gangrene’); see Blanc (2004) in <http://www.cairn.info/publications-de-Blanc-Alain--7916.htm>.

Jasanoff dismisses this so-called lex unda, proposing an ‘original’ -tino-, in his article <http://www.people.fas.harvard.edu/~jasanoff/pdf/Latin%20gerundive%20(preprint).pdf>. We would have then a group of passive participles-gerundives, which indicate possibility/obligation, built by agglutination of two suffixes; as, -ti-no-, -i-jo- (cf. O.Ind. ramanīya,  miśranīya), -tew-(ij)o-, -ti-mo- (cf. Gk. aidesimos), O.C.S. pečalьnъ  ‘deplorable’ (cf. O.Ir. fedelm), Lat.  amābilis, etc. Jasanoff also presents abstracts in -(n)ti-, which are behind the Latin gerund.

c. A future (or obligation) passive gerundive ending, -téw(ij)os, existed in Late Indo-European; as, legtéw(ij)os, which has to be said, read or gathered. Because of its passive use, it may be used only with transitive verbs.

NOTE.  For the absolutive use of -téw(ij)os, cf. Gk. -τος, -τεος, O.Ind. -tavya, O.Ir. -the, etc., probably all from verbal adjectives in -tu-, full grade -tew-, usually lengthened with common gerundive ending -ij-.

d. -m, with a general meaning of ‘able’; as, mnām, mindful.

NOTE. For the ‘internal derivation’ (after the German and Austrian schools) of this PIE suffix -m*-mon, cf. Gk.mnẽma < mń-m ‘reminder’, Gk. mnmon < mnām ‘who remembers’; compare also Skt. bráhman ‘prayer’, Skt. brahmán ‘brahman’, etc.

3. The adverbial, not inflected verbal adjectives are called absolutives or gerunds. They were usually derived from older gerundives.

NOTE. PIE speakers had to use verbal periphrases or other resources to express the idea of a modern gerund, as there were no common reconstructible PIE gerunds. Just like verbal nouns were the usual basis to express the idea of infinitives, verbal adjectives (and especially gerundives) were a common PIE starting point to create gerunds.

 


 

7.8. Conjugated Examples

7.8.1. Thematic Verbs

 I. Present Stem

Active

loutum, to wash (present stem low-o-)

 

Indicative

Subjunctive

Optative

Imperative

IMPERFECT

Sg.

lowō

lowō

lowoim

-

lowom

lówesi

lówēsi

lowois

lowe

lowes

lóweti

lówēti

lowoit

lówetōd

lowet

Pl.

womos

lówōmos

lówoime

-

lówomo

lówete

lówēte

lówoite

lówete

lówete

lówonti

lówōnti

lowoint

lówontōd

lowont

deiktum, to show (present stem deik-o-)

 

Indicative

Subjunctive

Optative

Imperative

IMPERFECT

Sg.

deikō

deikō

deikoim

-

deikom

déikesi

déikēsi

deikois

deike

deikes

déiketi

déikēti

deikoit

déiketōd

deiket

Pl.

déikomos

déikōmos

déikoime

-

déikome

déikete

déikēte

déikoite

déikete

deikete

déikonti

déikōnti

deikoint

déikontōd

déikont

weistum (<weid-tum), to see (present stem wid-jo-)

 

Indicative

Subjunctive

Optative

Imperative

IMPERFECT

Sg.

widējō

widējō

widējoim

-

widējom

widjesi

widjēsi

widējois

widēje

widējes

widjeti

widjēti

widējoit

widjetōd

widējet

Pl.

widjomos

widjōmos

widjoime

-

widjomo

widjete

widjēte

widjoite

widjete

widjete

widjonti

widjōnti

widējoint

widjontōd

widjont

 

Middle-Passive

loutum, to wash (present stem low-o-)

 

Indicative

Subjunctive

Optative

IMPERFECT

Sg.

lowar

lówōmar

lówoima

lowa

lówesor

lówēsor

lówoiso

lóweso

lówetor

lówētor

lówoito

lóweto

Pl.

lówomesdha

lówōmesdha

lówoimedha

lówomedha

lówedhwe

lówēdhwe

lówoidhwe

lówedhwe

lówontor

lówōntor

lówointo

lówonto

deiktum, to show (present stem deik-o-)

 

Indicative

Subjunctive

Optative

IMPERFECT

Sg.

deikar

déikōmar

déikoima

deika

déikesor

déikēsor

déikoiso

déikeso

déiketor

déikētor

déikoito

déiketo

Pl.

déikomesdha

déikōmesdha

déikoimedha

déikomedha

déikedhwe

déikēdhwe

déikoidhwe

déikedhwe

déikontor

déikōntor

déikointo

déikonto

weistum, to see (present stem wid-jo-)

 

Indicative

Subjunctive

Optative

IMPERFECT

Sg.

widējar

widjōmar

widjoima

widēja

widjesor

widjēsor

widjoiso

widjeso

widjetor

widjētor

widjoito

widjeto

Pl.

widjomesdha

widjōmesdha

widjoimedha

widjomedha

widjedhwe

widjēdhwe

widjoidhwe

widjedhwe

widjontor

widjōntor

widjointo

widjonto

 


 

II. Aorist Stem

Active

loutum, to wash (aorist stem sigmatic lou-s-)

 

Indicative

Subjunctive

Optative

Sg.

lous

lousō

lousīm

lous

lousesi

lousīs

loust

louseti

lousīt

Pl.

lousme

lóusomos

lóusīme

louste

lóusete

lóusīte

loust

lousonti

lousīnt

deiktum, to show (aorist stem dik-ó-, zero-grade)

 

Indicative

Subjunctive

Optative

Sg.

dikóm

dik

dikóim

dikés

diksi

dikóis

dikét

dikti

dikóit

Pl.

dikome

dikōmos

dikoime

dikete

dikēte

dikoite

dikónt

diknti

dikóint

NOTE. For original dikóm, cf. diśáti, Gk. ἄδικος, etc.

weistum, to see (aorist stem wid-ó-, zero-grade)

 

Indicative

Subjunctive

Optative

Sg.

widóm

wid

widóim

widés

widsi

widóis

widét

widti

widóit

Pl.

widome

widōmos

widoime

widete

widēte

widoite

widónt

widnti

widóint

NOTE. For PIE accent on the optative suffix, following the accent on the thematic vowel of certain Aorist formations, cf. O.Ind. them. aor. opt. sg. vidé-s (<*widói-s).

 

Middle-Passive

loutum, to wash (aorist stem lou-s-, sigmatic)

 

Indicative

Subjunctive

Optative

Sg.

lousma

lousar

lóusīma

louso

lóusesor

lóusīso

lousto

lóusetor

lóusīto

Pl.

lóusmedha

lóusomesdhā

lóusīmedha

lousdhwe

lóusedhwe

lóusīdhwe

lousto

lóusontor

lóusīnto

deiktum, to show (aorist stem dik-ó-, zero-grade)

 

Indicative

Subjunctive

Optative

Sg.

diká

dikr

dikoima

dikeso

dikēsor

dikóiso

diketo

dikētor

dikoito

Pl.

dikómedha

dikmesdhā

dikóimedha

dikedhwe

dikēdhwe

dikoidhwe

dikonto

dikōntor

dikointo

weistum, to see (aorist stem wid-ó-, zero-grade)

 

Indicative

Subjunctive

Optative

Sg.

widá

widr

widoima

wideso

widsor

widoiso

wideto

widētor

widoito

Pl.

widómedha

widmesdhā

widóimedha

widedhwe

widēdhwe

widoidhwe

widonto

widōntor

widointo

 


 

III. Perfect Stem

loutum, to wash (perfect stem lōu-/lou-)

 

Indicative

Subjunctive

Optative

Sg.

lōwa

lōwō

loujm

lōutha

lōwēsi

loujs

lōwe

lōwēti

loujt

Pl.

loumé

lwōmos

lowī

louté

lwēte

lowī

lowr

lōwōnti

lownt

 

deiktum, to show (perfect stem doik-/dik-)

 

Indicative

Subjunctive

Optative

Sg.

doika

doikō

dikjm

doiktha

doikesi

dikjs

doike

doiketi

dikjt

Pl.

dikmé

dóikomos

dikī

dikté

dóikete

dikī

dikḗr

doikonti

diknt

 

weistum, to see (perfect stem woid-/wid-, know)

 

Indicative

Subjunctive

Optative

Sg.

woida

woidō

widjm

woisthai

woidesi

widjs

woide

woideti

widjt

Pl.

widmé

wóidomos

widīme

wistéii

wóidete

widīte

widḗr

woidonti

widnt

From woid-tha.  ii From wid-.

 

7.8.2. Athematic Inflection

I. Present Stem

Active

estum, to be (present stem  es-/s-)

 

Indicative

Subjunctive

Optative

Imperative

IMPERFECT

Sg.

esmi

esō

sjēm

-

es

esi

ésesi

sjēs

es/sdhí

es

esti

éseti

sjēt

estōd

est

Pl.

smes

ésomes

sīme

-

sme

ste

ésete

sīte

(e)ste

ste

senti

esonti

sīnt

sentōd

sent

NOTE. Proto-Indo-European verb es-, be, exist, originally built only a durative aspect of present, and was therefore supported in some dialects (as Gmc., Sla., Lat.) by the root bheu-, be, exist, which helped to build some future and past formations.

 

kleutum, to hear (present stem kneu-/knu-, with nasal infix)

 

Indicative

Subjunctive

Optative

Imperative

IMPERF.

Sg.

kneumi

knewō

knujm

-

knew

kneusi

knéwesi

knujs

knéu/kḷnudhí

knéus

kneuti

knéweti

knujt

kneutōd

knéut

Pl.

knumés

knéwomos

knwīmé

-

knumé

knuté

knéwete

knwīté

knuté

knuté

knunti

knéwonti

knewīnt

klnwentōd

knúnt

NOTE. Indicative forms could have possibly been read with a columnar accent in a post-Late Indo-European period, i.e. knumes, knute, opt. knwīme, knwīte, imp. knudhi, as in Greek.

The 3rd pl. optative had full-grade root vowel, see Kortlandt (1992), Beekes (1995).

 

 

stātum, to stand (present stem (si)stā-/(si)sta-, reduplicated)

 

Indicative

Subjunctive

Optative

Imperative

IMPERFECT

Sg.

()stāmi

stájō

(si)stajm

-

(si)stām

()stāsi

stájesi

(si)stajs

stā/stadhí

(si)stās

()stāti

stájeti

(si)stajt

stātōd

(si)stāt

Pl.

(si)stamés

stájomos

(si)staīmé

-

(si)sta

(si)staté

stájete

(si)staīté

staté

(si)sta

(si)stanti

stájonti

(si)stant

stanti

(si)stant

NOTE. Indicative forms were possibly read in post-LIE period with columnar accent, as sístames, sístate, etc. or stames, state, etc.

The optative formations show zero-grade stem sta-, and the accent is written to distinguish -a-ī- from a diphthong --.

For sta-jo- as a thematic subjunctive (cf. O. Gk. subj. 1st pl. στεομεν (< PGk stejo- < LIE *stəjo- > NWIE stajo-), from Gk. στημι; also, θεομεν (<PGk dhejo- < LIE *dhəjo- > NWIE dhajo-) from τθημι, IE dhē-; δεομεν (<PGk dejo- < LIE *dəjo- > NWIE dajo-) from Gk. δίδωμι, IE dō; and so on.

 

Middle-Passive

kleutum, to hear (present stem kneu-/knu-, with nasal infix)

 

Indicative

Subjunctive

Optative

IMPERFECT

Sg.

knumár

knéwomar

knwīmá

knumá

knusór

knéwesor

knwīsó

knusó

knutór

knéwetor

knwītó

knutó

Pl.

knumesdha

knéwomesdha

knwīmedha

knumédha

knudhwé

knéwedhwe

knwīdhwé

knudhwé

knuntór

knéwontor

knwīn

knun

NOTE. In a post-LIE period a common columnar accent would have been also possible; viz. knumar, knusor, etc.


 

stātum, to stand (present stem (si)stā-/(si)sta-, reduplicated)

 

Indicative

Subjunctive

Optative

IMPERFECT

Sg.

(si)stamár

stájomar

(si)staīmá

(si)stama

(si)stasór

stájesor

(si)staīsó

(si)staso

(si)statór

stájetor

(si)staītó

(si)stato

Pl.

(si)stamesdha

stájomesdha

(si)staīmedha

(si)stámedha

(si)stadhwé

stájedhwe

(si)staīdhwé

(si)stadhwe

(si)stantor

stájontor

(si)staīn

(si)stanto

 

II. Aorist Stem

Active

kleutum, to hear (aorist klew-/klu-)

 

Indicative

Subjunctive

Optative

Sg.

klewom

klewō

klujm

klewes

kléwēsi

klujs

klewet

kléwēti

klujt

Pl.

klwome

kléwōmos

klwī

klwete

kléwēte

klwī

klwont

kléwōnti

klwīnt

NOTE. For aorist klew-/klu- cf. Gk. ἔ-κλυον, O.Ind. áśrot.

stātum, to stand (aorist stem stā-)

 

Indicative

Subjunctive

Optative

Sg.

stām

stajō

stajm

stās

stajesi

stajs

stāt

stajeti

stajt

Pl.

sta

stájomos

staī

sta

stájete

staī

stnt

stajonti

staīnt

 


 

Middle-Passive

kleutum, to hear (aorist stem kluw-)

 

Indicative

Subjunctive

Optative

Sg.

klwá

kléwomar

klwī

klwesó

kléwēsor

klwīsó

klwetó

kléwētor

klwītó

Pl.

klwomesdha

kléwōmesdha

klwīmedha

klwedhwé

kléwēdhwe

klwīdhwé

klwontó

kléwōntor

klwīn

stātum, to stand (aorist stem stā-)

 

Indicative

Subjunctive

Optative

Sg.

stamá

stajar

staīmá

stasó

stájesor

staī

stató

stájetor

staītó

Pl.

stamedha

stájomesdha

staīmedha

stadhwé

stájedhwe

staīdhwé

stantó

stájontor

staīn

 

III. Perfect Stem

kleutum, to hear (perfect stem -klou-/-klu-, reduplicated)

 

Indicative

Subjunctive

Optative

Sg.

kéklowa

kéklowō

keklujm

kékloutha

kéklowesi

keklujs

kéklowe

kékloweti

keklujt

Pl.

keklumé

kéklowomos

keklwīmé

kekluté

kéklowete

keklwīté

keklwḗr

kéklowonti

keklwnt

 

 

 

stātum, to stand (perfect stem se-stā-/se-sta-, reduplicated)

 

Indicative

Subjunctive

Optative

Sg.

sestā

séstā

sestajm

séstātha

séstājesi

sestajs

sestā

séstājeti

sestajt

Pl.

sestamé

séstājomos

sestaīmé

sestaté

séstājete

sestaīté

sestr

séstājonti

sestant

NOTE. For reduplicated s(t)e-sta-, cf. O.Ind. perf. tastháu, Av. -šastarə, Gk. héstamen, O.Ir. -sestar.

7.8.3. Common PIE Stems

I. Thematic Verbs

1. Root:

o     Present lowō, I wash, aorist lous, perfect lélowa.

o     Present serpō, I crawl, aorist spóm.

o     Present bherō, I carry, aorist bhērm, perfect bhébhora.

o     Present bheugō, I flee, aorist bhugóm.

o     Present bheidhō, I believe, persuade, aorist bhidhom.

o     Present weqō, I speak, aorist (them. redupl.) weuqom.

o     Present tremō, I tremble, aorist tmom.

NOTE. A particular sub-class of thematic presents without suffix is of the type Skt. tudáti, which have present stems with zero-grade root-vowel, as glubhō/gleubhō, skin.

2. Reduplicated:

There are many reduplicated thematic stems, analogous to the athematic ones:

o     Present gignō, I beget, (from gen-), aorist gnom/genóm, perfect gégona, p.part. gnātós.

NOTE. For gnātós, cf. O.Ind. jātás, Av. zāta-; Lat. nātus, Pael. cnatois, Gaul. f. gnāthadaughter”; O.N. kundrson”, also in compound, cf. Goth. -kunds, “ be a descendant of “, O.E. -kund, O.N. -kunnr.

o     Present pibō, I drink (<reduplicated -, from pōi-).

o     Present mimnō, I remember, (from men-).

3. In -je/o-, some of them are causatives:

o     Present spekjō, I watch, aorist speks, p.part. spektós.

o     Present tenjō, I stretch, aorist tnom/tenóm, perfect tétona, p.part. ttós.

4. Verba vocalia:

o     Present widējō, I see, I know, aorist widóm, perfect woida p.part. wistós (<wid-tós).

o     Present monejō, I make think, warn, as Lat. moneo, from men-, think.

o     Present tromejō, I make tremble, from trem-, tremble.

5. In -ske/o-:

o      Present pksk, I ask, demand, inquire (cf. Lat. posco, Ger. forschen) from prek-, ask.

o     Present gnāskar, I am born (cf. Lat. gnascor), p.part. gnātós, from zero-grade of gnō-, beget.

o     Present ()gnoskō, I begin to know, I learn, from  gnō-, know.

6. With nasal infix:

o     Present jungō, join (from jeug-), aorist jeugom; p.part. jugtós.

NOTE. Compare O.H.G. (untar-)jauhta (as Lat. sub-jugaui), Lat. jungō, -ere, -nxi, -nctus, Gk. ζεύγνῡμι, ζεῦξαι ζυγηναι; O.Ind. yunákti (3. Pl. yuñjánti = Lat. jungunt), yuñjati, full-grade yōjayati (<jeugējeti); Av. yaoj-, yuj-; Lit. jùngiu, jùngti, etc. For past participles (with and without present infix -n-), compare O.E. geoht, iukt, Lat. junctus, Gk. δεπθηόο, O.Ind. yuktá-, Av. yuxta-, Lit. jùngtas, etc.

II. Athematic Verbs

1. Root: They are the most archaic PIE verbs, and their present conjugation is of the old type singular root vowel in full-grade, plural root vowel in zero-grade.

o      Present esmi, I am.

o      Present eimi, I go.

o      Present bhāmi, I speak.

NOTE. The verb talk is sometimes reconstructed as PIE *āmi, I talk, andimperfect *ām, I talked/have talked; for evidence of an original ag(h)-, compare Lat. aiō, Gk. ην, Umb. aiu, Arm. asem. Thus, this paradigm would rather be thematic, i.e. present ag(h), I talk, vs. imperfect ag(h)jóm, I talked/have talked.

o      Present edmi, I eat.

NOTE. Note that its early present participle dent-, “eating”, was frozen as substantive dentis, meaning “tooth”.

o      Present welmi, I want.

2. Reduplicated:

o     Present ()stāmi (from stā-, stand), aorist stām, p.part. statós.

o     Present déidikmi (from deik-, show), aorist dēiks, perfect dédoika, p.part. diktós.

o     Present dhédhēmi (from dhē-, do, make), aorist dhēm, p.part. dhatós.

o     Present dídōmi (from -, give), aorist dōm, p.part. datós.

o     Present jíjēmi, expel, aorist jem.

NOTE. For evidence on an original PIE jíjēmi, and not ˟jíjāmi as it is sometimes reconstructed, cf. Lat. pret. iēcī, a form due to its two consecutive laryngeals, while Lat. iaciō is a present remade (J. González Fernández, 1981).

3. With nasal infix:

o      knéumi, hear (from kleu-), aorist klewom, perfect kéklowa, p.part. klutós, heard, also ‘famous’.

o      punémi, rot (from pew), aorist pēws.

7.9. Verbal Composition

Verbs were often combined with adverbs to modify their meaning. Such adverbs were called preverbs and in the first instance remained separate words. Over time they tended to join with verbs as prefixes.

NOTE. For more on preverbs, see below §10.6.1.

Occasionally, verbs were compounded with a non-adverbial element, such as a noun. The most familiar example of this is kréd-dhē-mi, believe, trust, literally ‘place one’s heart in’, cf. Vedic Skt. śrád dadhāti, Lat. crēdō, and O.Ir. cretim (Fortson 2004).

7.10. The Verbal Accent

The finite verb of a LIE main clause was normally placed following the subject and the object, at the end of the sentence, where the sentence accent usually decreases. However, when the verb was stressed at the beginning of the sentence, or in a subordinate clause, it carried its normal accent.

NOTE. Fortson (2004): “In Vedic Sanskrit, main-clause finite verbs that do not stand at the beginning of their clause are written in the manuscripts without accent marks. In Greek, the rules for accenting verbs are different from those for nouns, and resemble the accentuation of strings of clitics; this suggests an affinity between the prosody of verbs and the prosody of chains of weakly stressed or unstressed particles. In Germanic heroic poetry, fully stressed words alliterate with one another, but certain verbs, together with unstressed pronouns and particles, do not participate in alliteration; this suggests weaker prosodic status for those verbs. In certain Germanic languages, such as modem German, verbs are required to be the second syntactic unit in main clauses, which is the same position taken by many unstressed sentence particles elsewhere in Indo-European (Wackernagel’s Law)”.

Meier-Brügger (2003) also states that “[r]esearchers agree that Vedic generally reflects the fundamental characteristics of Proto-Indo-European, and thus, that the finite verb in a main clause was unstressed (…) It remains disputed whether the second position of the finite verb, common to modern Germanic languages such as German, originated from the inherited phenomenon of enclitics, or whether it appeared secondarily”. On that, Wackernagel (1892) “the German rule of word order was already valid in the mother language”.

Finite verbs were therefore prosodically deficient in PIE, they could behave as clitics, i.e. they had no stress and formed an accentual unit with a neighbouring stressed word. However, they were fully stressed when moved to the front of a clause for emphasis or contrast, or when occurring in subordinate clauses. See below §10 Syntax.