Pinkster, Harm (1942-) [1990], Latin Syntax and Semantics [info], xii, 320 p.: ill.; 24 cm. [word count] [Pinkster].
Previous SubSect

Next SubSect

11.1. Definitions 11.1.1 States of affairs and `Aktionsart'

In crosssection 2.4. we have encountered `dynamic' and `non-dynamic' states of affairs. Examples are (1) and (2), respectively (cf. (48) and (55) in crosssection 2.4.):

(1) the boy got a book

(2) the boy had a book

(1) describes a change (the transition from a state of `not-having' to one of `having'); (2), on the other hand, describes a state. With the aid of the distinction `dynamic' (example (1)) vs. `non-dynamic' (example (2)) we indicate how the state of affairs expressed by the nuclear predication evolves in time.

In reality, a finer distinction can be made within the category of dynamic states of affairs, viz. between `terminative' and `non-terminative' states of affairs, cf. (3)–(4):

(3) the boy wrote a book

(4) the boy read a book

[1a] (3) concerns an action which, unless it is interrupted, naturally results in a finished book (a book in this example is often called an `effected Object'). In (4) the action can be finished, just as in (3), but does not involve a natural end product. The difference between (3) and (4) can be brought to light by adding

-- 215 --

Time Adjuncts. `In three hours' can be added to (3), `for three hours' cannot; this does not apply to (4).

(3') the boy wrote a book in three hours/* for three hours

(4') the boy read a book? in three hours [2]/for three hours

The category of terminative states of affairs can be subdivided into `momentaneous' and `non-momentaneous' states of affairs, see (5) and (3), respectively:

(5) the boy found a book

(3) the boy wrote a book

Whereas (3) can be interrupted, (5) cannot:

(5') *the boy stopped finding a book

(3") the boy stopped writing a book [3]

These distinctions between types of states of affairs can be schematized as in figure 11.1: [4]

Latin examples of the distinctions made here are:

(6) duodequadraginta annos tyrannus Syracusanorum fuit Dionysius, cum … (`D. had been ruler of S. for 38 years, when …', Cic. Tusc. 5.57) [(a): –dyn.]

(7) cum iam tot horas de uno genere … dicam (`Now that I have been speaking about one subject for so many hours already', Cic. Ver. 5.159) [(b):-term.]

-- 216 --

(8) ut una hora … reo … spem … praeciderem (`That within one hour I robbed the accused of the hope …', Cic. Ver. 1.20) [(c):-mom.]

(9) ipse … Tarraconem paucis diebus pervenit (`He himself reached T. in a few days', Caes. Civ. 2.21.4) [(d): + mom.]

In all honesty I should remark that it is not always easy to distinguish between non-dynamic and non-terminative states of affairs in Latin. In English the difference appears from the fact that non-dynamic states of affairs cannot very well be used in the progressive (* I was having a book), whereas to read a book does allow this construction. Furthermore, adverbs indicating the tempo (e. g. celeriter (`rapidly') or another aspect of the manner in which the action takes place (e. g. diligenter (`diligently') cannot very well be added to non-dynamic states of affairs. Yet, these are not absolute rules.

In Latin linguistics little attention has been devoted to the classification of states of affairs on the basis of well-defined criteria (but see note 4). The grammars (e. g. Sz. 303 ff.) do offer a more or less intuitive subdivision of verbs (rather than states of affairs!) into verbs with an `imperfective' `Aktionsart' and verbs with a perfective `Aktionsart'. [5] In this connection, `imperfective' roughly coincides with non-dynamic and non-terminative in figure 11.1, i. e. the division largely coincides with the addibility of Time-during-which Adjuncts and Time-within-which Adjuncts, respectively. An important difference between the approach of Sz. (and others) and that outlined above lies in the fact that the former concerns the Aktionsart of verbs and the latter that of states of affairs. In crosssection 3.1. (p. 26) it has already been pointed out that the addibility of Adjuncts like those mentioned here does not merely depend on the meaning of the verb, but also on properties of the arguments required by the verb. See (3') and (10), with a singular and a plural Object, respectively:

(3') *the boy wrote a book for three hours

(10) the boy wrote rhymes for three hours

Another significant factor is negation. In states of affairs with the verb venire (`to come', with or without a Direction Adjunct) we do not as a rule find Duration Adjuncts, but see (11):

(11) Philippus nullus usquam nec nuntius ab eo per aliquot horas veniebat (`P. was nowhere in sight, and for several hours no messenger arrived from him', Liv. 32.35.2)

While `to come from someone to something' is terminative, `not to come' is, in turn, non-dynamic and thus allows a Duration Adjunct. The verb alone, therefore, is not a satisfactory starting-point.

In Latin morphology there are means to create verbs on the basis of a verbal stem (e.g. derivation); the resulting verbs differ from the verbum simplex in meaning and, together with the arguments they govern, indicate a different state of affairs. Examples are the formation of so-called inchoatives from

-- 217 --

non-dynamic verbs by means of the infix -sc- (e. g. calere → calescere (`to be/become warm')), and the formation of causatives by means of composition with facere (e.g. calere → calefacere (`to be/make warm')). These verbs occur in different states of affairs: non-dynamic (calere) vs. terminative (calescere, calefacere). Similar differences may be seen between active and passive. A fourth example is offered by compounds formed by means of a preverbial prefix, e.g. edere → comedere (`to eat/to consume'), suere → consuere (`to sew/to stitch together'). Preverbial prefixes often turn a one-place verb (or a two-place verb with a `reduced' one-place counterpart, see p. 8) into a two-place verb (or a non-reducible two-place verb) (in traditional terms: they turn intransitives into transitives). This is often accompanied by a difference in the type of state of affairs of the nuclear predication in which these verbs occur (non-dynamic → terminative) (see Sz. 287; 304). [6]

Much attention has been devoted in Latin linguistics to the distinction between the so-called perfective (i.e. terminative) and imperfective (i.e. non-dynamic and non-terminative) verbs, without, however, producing many reliable results. Sz. (303), for instance, remarks that of the so-called `perfective' verbs (i. e. terminative states of affairs) the imperfect as a rule does not occur and that, if it does occur, such an imperfect is to be interpreted as iterative or as `conative'. I return to this in crosssection 11.2.1.1.4 on p. 227. Generally speaking, it has been observed that verbs of certain semantic classes predominantly occur in certain tenses. This is, of course, correlated with the fact that owing to their meaning verbs are especially suited for certain contexts: verbs meaning `to consider', for instance, often (but not always) present background information to certain main actions and in narrative texts, therefore, often occur in the imperfect (see crosssection 11.2.1.1.4). [7] See (12):

(12) sic primum existimabam … Simul et verebar … Itaque … eram, … exquirebam … cognoscebam … revertebar … explicabam … exspectabam … erat in animo. Cum haec agerem, repente ad me venit Heraclius (`First I thought … At the same time I was afraid … So I was …, I enquired …, I tried to find out …, I returned to …, I explained …, I expect ed …, I was planning … While I was busy with all this, suddenly H. came to me', Cic. Ver. 4.136–7)

11.1.2 Tense

`Tense' is a `deictic' category, by which the state of affairs expressed by the predication is located in time. The state of affairs can either be related to the moment of writing or speaking, or to another moment known from context or situation. Thus, laudo as a rule means that the action of praising is contemporaneous with the moment of speaking (`I am (now) praising'), laudaveram that the praising was anterior to a moment that itself is anterior to the moment of speaking (`I had praised at a moment that precedes the present

-- 218 --

moment'). In other words, tense is a morpho-semantic category which allows the chronological ordering of events. Besides marking of the verb form, also lexical means can be used to indicate this chronological order of states of affairs, such as temporal adverbs (heri (`yesterday'), nunc (`now'), postea (`afterwards'), deinde (`then'))-in a number of languages the only possibility-or simply by the order in which the events are mentioned, e. g. (13):

(13) veni, vidi, vici (`(first) I came, (then) I saw and (after that) I conquered', Suet. Jul. 37)

In this chapter I only deal with tense as a verbal category.

In locating a certain state of affairs in time (and, thus, in choosing a verb form), the language user is as a rule bound by the actual chronological relation between events. This appears, for instance, from obvious impossibilities such as (14):

(14) *heri laudabo (*`Yesterday I shall praise')

On the other hand, it is not true that the language user has no choice at all. Thus, in Latin, as in English, the imperfect can be used also if there is no cogent temporal reason for this, e.g. example (15):

(15) sed si domi est, Demaenetum volebam (`But I wanted to see D., if he is at home', Pl. As. 452)

By locating his request in the past, the speaker in this example presents this request as less urgent or immediate, and thus as slightly more modest or polite. Tense can, therefore, to some extent be used, like the category mood, to express the attitude of the speaker. [8]

With the aid of the examples laudo and laudaveram I have shown that a specific tense can be related either to the moment of speaking (the `present' moment) or to another moment that itself is related to the moment of speaking. In some grammars the former (laudo) is called `absolute' use of the tense, the latter (laudaveram) `relative' use of the tense. In reality, both are relative, but the latter has, as it were, two stages. The Latin verb turns out to be construed systematically according to two dimensions:

(i) most verb forms contain information as to the chronological order (anterior, contemporaneous, posterior) of the predication [9] with regard to a past, present or future moment known from context or situation;

(ii) part of the verb forms, especially indicative forms, also contain information (apart from that mentioned in (i)) as to the location of the predication in time (past, present, future).

Note that in subjunctive forms only present and past are distinguished. They are left out of account here.

Examples of the two dimensions mentioned above are:

-- 219 --


(i) (a) Participles indicate whether the (embedded) predication expressed by them is anterior, contemporaneous or posterior with regard to another predication (as a rule the main predication): locutus/loquens/locuturus (`after speaking/while speaking/before speaking'). The participles do not indicate when the action in question occurred: in loquens abiit (`While speaking he left') the speaking occurred in the past; in loquens abibit (`while speaking he will leave') the speaking will occur at the future moment of leaving.

(b) In a similar way, infinitives indicate anteriority, contemporaneous|ness or posteriority of the predication with regard to the (main) predication: Cicero dicit Caesarem in senatu locutum esse/loqui/locuturum esse (`Cicero says that Caesar has spoken/is speaking/will speak in the senate').

(c) Finite verb forms of the perfectum stem as a rule [10] indicate that the predication has occurred before a moment known from context or situation, see above laudaveram and (16):

(16) rogo quid fecerit (`I ask what he has done')

N.B.: see also (ii) (a) below.

(d) Finite verb forms of the infectum stem indicate that the predication occurs contemporaneously with a moment known from context or situation, see above laudo and (17):

(17) rogavi quid faceret (`I asked what he was doing')

N.B.: see also (ii) (b) below.

(ii) (a) Perfect, pluperfect and future perfect locate a predication anterior to present, past and future, respectively. Caesar locutus erat (`Caesar had spoken') means that Caesar's speaking preceded a certain moment in the past.

(b) Present, imperfect and (simple) future locate a predication in present, past and future, respectively; on the basis of what was said under (i) (d) e.g. Caesar loquebatur means that at a certain moment in the past Caesar was speaking.

(c) The periphrastic forms locuturus sum/eram/ero offer a limited possibility to locate predications posterior to present, past and future, respectively.

Tables 11.1 and 11.2 offer a schematic representation of what has been remarked above. In a different form, table 11.2 returns in the discussion of Table 11.1 Non-finite verb forms (participles, infinitives)
AnteriorContemporaneousPosterior
locutusloquenslocuturus
locutus esseloquilocuturus esse

-- 220 --

Table 11.2 Finite verb forms (1st person singular indicative)
Location with regard to orientation moment
Orientation momentAnteriorContemporaneousPosterior
Presentlocutus sumloquorlocuturus sum
Pastlocutus eramloquebarlocuturus eram
Futurelocutus eroloquarlocuturus ero
Table 11.3 Sequence of tenses
Tense of main clauseTense of subordinate clause
AnteriorContemporaneousPosterior
Present Futureperfectpresentperiphrastic
future
(locuturus sim)
Pastpluperfectimperfectperiphrastic
future
(locuturus essem)
the rules concerning the correspondence between the tenses of main and subordinate clauses in the subjunctive governed by them (the sequence of tenses (consecutio temporum)). See table 11.3. From table 11.2 it may be deduced that the system apparently contains some elements of redundancy. What is, for instance, the de facto difference between anterior to the present (locutus sum) and contemporaneous with the past (loquebar)? After all, both forms concern past events. Likewise: what is the difference between loquar and locuturus sum (two forms that in reality are not often distinguished, as has been pointed out above)? I return to this in crosssection 11.2. on p. 223 ff. [11]

11.1.3 Aspect

The notion `aspect' appeared in the second half of the nineteenth century, first in Greek and later also in Latin linguistics, following the use of the notion `vid' in Slavic linguistics, where it had been introduced owing to the inadequacy of the category `tense'. It served, among other things, to explain differences between pairs of non-compound and compound verbs that may ostensibly be used to refer to the same actions, e.g. the paradigms of the Russian verbs meaning `to write' in table 11.4. The difference between these verbs was described as `in progress' (the left-hand column) vs. `gone by' (the right-hand column), or as `not (yet) completed' vs. `completed' (`imperfective' vs. `perfective'). Assuming this difference, one can also explain why the form ya napishú is not a present tense (comparable to ya pishú): a `gone by', `completed' present is difficult to imagine. Practically the form functions as a future.

-- 221 --

Table 11.4
Act. infinitivepisát'napisát'(`to write')
1st pers. sing imperf.ya pisállya napisáll(`I wrote')
Pres.ya pishú-(`I write')
Fut.ya búdu pisát'ya napishú(`I shall write')

The notions `aspect', `imperfective' and `perfective' also seemed useful for describing the semantic difference between the Latin perfect and imperfect. In this connection, scholars also pointed at the terminology used by the Greek and Latin grammarians, e.g. the Latin terms `praeteritum perfectum' and `praeteritum imperfectum'. `Perfective aspect' was ascribed to the perfect, `imperfective aspect' to the imperfect. We see that the notion `aspect', mainly used in Slavic grammar to explain differences between pairs of verbs, was introduced into Greek and Latin grammar in order to explain differences between forms of one and the same verb.

N.B.: In the preceding section I have stated that the terms `imperfective' and `perfective' are also used to describe properties of verbs. I return to this in crosssection 11.1.4 on p. 223.

The question of whether or not in Latin `aspect' `exists' or is necessary for the description of the uses of the various verb forms has led to a flood of scholarly publications. The communis opinio is that in an earlier stage of Latin aspect was probably or certainly a productive category, but that the Latin of the texts we have at our disposal contains few (or no) traces of this. In Classical Latin contemporaneousness and anteriority are the most important factors. There is a rather persistent tendency to explain statistically remarkable, idiomatic uses of certain verb forms as relics of an earlier aspectual system. An example of such an expression is (18):

(18) cuius ego iudicium, pace tua dixerim, longe antepono tuo (`Whose judgment, if I may say so with your permission, I value much more than yours', Cic. Tusc. 5.12)

In instances of this kind, dixerim cannot be taken as anterior to the present. More of these instances will be discussed below.

One explanation for the popularity of the aspectual view of the Latin tense system might be the unspoken wish to find in Latin a system similar to that found in Greek. Yet, the usefulness of the category `aspect' for Greek is by no means undisputed. [12] Moreover, an important difference between Latin and Greek is the fact that Greek almost always offers the language user a `choice' between pairs of verb forms. In Latin this is not the case (apart from the perfect and imperfect indicatives and some idiomatic uses). Cf. the Greek examples (19a–b) and (20a–b) of present and aorist imperatives and infinitives, respectively: [13]

-- 222 --


(19a) ἀναγνώσεται πρῶτον ὑμῖν τὸν τούτου νόμον, εἶτα τοὺς ἄλλους ... ἀναγίγνωσκε (`First he will read to you his law, then the other laws, … Please read', Dem. 24.39)

(19b) ἀνάγνωθι δὲ καὶ τούτους τοὺς νόμους (`And read those laws as well', Dem. 24.104)

(20a) … . . . οἱ μάγοι ἀνεβίβασαν αὐτὸν ἐπὶ πύργον καὶ ἀγορεύειν ἐκέλευον (`The Magi brought him up on to a tower and ordered him to speak', Hdt. 3.75.1)

(20b) κεῖνον . . . ἐκέλευον ἀναβάντα ἐπὶ πύργον ἀγορεῦσαι ὡς … (`They ordered him to go up on to a tower and declare that … ', Hdt. 3.74.3)

Another explanation may be found in the fact that it is not made very clear what exactly `aspect' is. It is usually described as different from tense, e.g. as a non-deictic temporal category covering phenomena such as `completeness', `iteration', etc. In Latin linguistics the notions `progress' and `completeness' are most often used. I will now further discuss these two notions.

The grammarian Priscian (sixth century AD) describes the value of the Latin imperfect and perfect as follows:

(21) si incipiam in praeterito versum scribere et imperfectum eum (scil. versum) relinquam tunc utor praeterito imperfecto dicens `scribebam versum'; … continuo enim scripto ad finem versu dico `scripsi versum' (`If in the past I begin writing a verse and I leave it unfinished, I use the praeteritum imperfectum and say "scribebam versum"; … for immediately after writing the entire verse I say "scripsi versum", Prisc. In G.L. 8. 52–3)

In this example of the perfect the notion `complete', `finished' is entirely clear. The example is, however, not representative. We are dealing with a `terminative' state of affairs and in such cases a `result' interpretation of the perfect is obvious. With other Object constituents with scribere or with other verbs this interpretation is less obvious, see (22) and (23), respectively:

(22) nostri praeceptores servum cervumque u et o litteris scripserunt (`My teachers wrote servus and cervus with uo' (i. e. seruos and ceruos), Quint. Inst. 1.7.26)

(23) tun heri hunc salutavisti? (`Have you greeted him yesterday?', Pl. Am. 717)

Examples (21)–(23) have in common that before the moment of speaking the state of affairs has `ended', is `over' or at least that the statement of the language user refers to a phase of a situation or event that has at that moment been finished. See also (24):

(24) satis … vixi: invictus enim morior (`I have lived long enough; for I die without having been defeated', Nep. Ep. 9.4)

-- 223 --

Here the speaker is not dead, but is speaking of the stretch of life he has lived until then (*satis vivebam would be absolutely impossible). (See further also examples (69)–(71) in crosssection 11.3.1. on p. 238). Evidence for the `endedness' of a state of affairs marked by a perfect form may be found in the distribution of quantifying expressions such as semel (`once'), bis (`twice'), ter (`thrice'). These occur with the perfect, not with the imperfect, e.g. (25a–b):

(25a) sed fui … bis in Bithynia (`But I have twice been in Bithynia', Cic. Planc. 84)

(25b * sed eram bis in Bithynia

Quantifying expressions of this kind can typically be applied in retrospect to ended states of affairs (see further crosssection 11.2.1.1.5 on p. 230).

Thus, inasmuch as Latin linguists have used the ancient grammarians' notion `endedness' to support an aspect theory, it is not clear what exactly is the difference between the notion `completeness' and the notion `anteriority' as mentioned in crosssection 11.1.2 above. To my mind, the difference is solely terminological. I conclude, therefore, that it would be better from a methodological point of view to explain as much as possible the various uses of the individual verb forms with the aid of the two-dimensional description of crosssection 11.1.2 rather than to use the notion `aspect' in isolated cases.

11.1.4 A terminological digression

I have already pointed out (p. 217; p. 221) that in Latin linguistics the terms `imperfective' and `perfective' are often used for semantic differences between verbs as well as for semantic differences between verb forms, i.e. both for Aktionsart (more or less equivalent to state of affairs) and for aspect. In reality, the confusion is even greater. In many German publications the notion Aktionsart is used to indicate what I have called Aktionsart and aspect; in French works, on the other hand, these two notions are covered by the single term `aspect'. In English studies `lexical aspect' (Aktionsart) and `grammatical aspect' (aspect) are sometimes distinguished, but often the term `aspect' is used indiscriminately.

Previous SubSect

Next SubSect


Pinkster, Harm (1942-) [1990], Latin Syntax and Semantics [info], xii, 320 p.: ill.; 24 cm. [word count] [Pinkster].
Powered by PhiloLogic