Pashto verbs: tenses, aspect and the ergative past

At a glance

• Every Pashto verb has a present stem and a past stem, and each comes in two aspects, imperfective and perfective. Grammars of Pashto build their tables on these four forms. [2, 3]

• Most verbs make the perfective by putting a stressed wə́- (و) in front of the stem. Prefixed and compound verbs make it by moving the stress instead. [2, 3]

• Pashto has no tense of its own for the future. The little word bə (به) with a perfective verb carries it. [2, 3]

• In the past tense of a transitive verb, the one who acts takes the oblique case and the verb agrees with the thing acted on. In every other sentence the verb agrees with the subject. [1–3]

• Small words such as bə (به) and the weak pronouns me (مې), de (دي) and ye (يې) do not stand where their meaning puts them. They follow the first stressed element of the clause. [2–4]

• Grammarians do not yet agree on the names of the tenses, or on whether "perfective" is the right word. [2, 3]

The two stems and the ergative past are introduced in the alphabet and grammar article; what follows goes further into the verb. Dialect differences in sound are in Pashto dialects.

Two stems, four forms

Raverty, a British officer whose grammar of Pashto reached its third edition in 1867, wrote that the inflections of Afghan verbs are formed "from two original tenses only", the past and the aorist, which he names with their Arabic terms. [1] Later grammars still start from two stems. Habibullah Tegey and Barbara Robson, whose Reference Grammar of Pashto (1996) describes the central dialect, say every verb construction uses one of four forms: present imperfective, present perfective, past imperfective and past perfective. [2]

Anne Boyle David, whose 2014 grammar covers General Pashto and the Waziri and Dzadrani dialects, calls the four forms the bases of the verb. She notes, with Grjunberg (1987), that Pashto verbal morphology is unusually complex beside the verbs of other Iranian languages. [3] Tegey and Robson set out the four forms for the verb taṛəl (تړل), "to tie". [2]

• present imperfective: taṛ- (تړ)

• present perfective: wə́taṛ- (وتړ)

• past imperfective: taṛə́l- (تړل)

• past perfective: wə́taṛəl- (وتړل)

Regular verbs make the past stem by adding the stressed suffix -əl, which is also the ending of the infinitive. [2, 4] Personal endings go on the stem. In the present, the third person ending is the same for "he" and "she". In the past, the third-person endings change with gender. David lists gender among the categories of the verb only for third-person past forms and for the third person singular of "to be". [3, 4]

Some verbs have a past stem that differs from the present stem. Tegey and Robson call them irregular, and call a verb doubly irregular when the perfective and imperfective stems differ as well. David calls verbs with more than one stem "strong" and verbs with a single stem "weak". [2, 3]

The prefix wə- and the stress

Tegey and Robson say simple verbs form the perfective stem by adding the prefix wə́- to the imperfective stem. [2] David calls these the first conjugation. Their prefix carries its own stress, and the syllables after it are unstressed. [3] Tegey and Robson show wə́- taking heavy stress, like the negative nə́ (نه), and give forms of "stand": [2]

درېدم

daredə́m

"I was standing"

ودرېدم

wə́daredəm

"I stood"

The same grammar gives khwri (خوري), "he is eating", and nə́ khwri (نه خوري), "he is not eating", where the negative takes the stress. [2]

Not every verb takes wə-. David's second conjugation is the prefixed verbs, such as dar-kawəl "to give to you"; they form the perfective by shifting the stress to the prefix. His third conjugation is the verbs made from a noun or adjective plus a helping verb. They shift the stress onto the first element and use an irregular perfective form of the helping verb. [3] Tegey and Robson describe a group, the doubly irregular verbs, in which stress is the only difference. For "to plant" the present imperfective is kenaw- and the present perfective kénaw-, with the stress on the first syllable; ordinary spelling does not show it. [2]

What the perfective means is not settled. Tegey and Robson say it appears to relate to whether the action is completed, but add that the distinction is not consistent. Positive commands use the perfective and the matching negative commands use the imperfective. They conclude that it is best treated as a difference of grammatical form only, and warn readers not to match it with the English perfect. [2] David goes further and replaces the terms. She calls the two aspects "continuous" and "aorist". Her reasons are that "perfective" suggests completion, that it invites confusion with the perfect tenses, and that the twentieth-century Pashto grammarian Khan (2002) used "aorist", which she says fits the traditional Pashto terms mutlaq (مطلق) and mujarrad (مجرد) as Penzl reported them. [3] The more common names, imperfective and perfective, are used here.

Simple and compound verbs

A simple verb has one stem, like taṛəl. A compound verb joins a noun or adjective to a helping verb: kawəl, "to do, make", for a transitive verb, and kedəl, "to become", for an intransitive one. Tegey and Robson call them derivative verbs. David uses the wider term complex verbs, which also takes in verbs with separable prefixes, and says new verbs enter Pashto through noun or adjective plus verb. [2, 3] Taylor Roberts, whose sentences come from Yusufzai and Kandahari speakers, calls compound verbs a fairly open class that makes up the majority of verbs in the language. [4]

The adjective joṛ (جوړ), "healthy, constructed", gives joṛégəl, "to get better, be sewn, be built", and joṛawəl, "to build, make healthy, sew". [2] Tegey and Robson print two sentences with it: [2]

ليلا به جوړه شي.

layla bə jóṛa si

"Layla will get better."

اسد به دېوال جوړ کړي.

asad bə dewāl jóṛ kṛi

"Asad will build a wall."

Compound verbs have a split character that grammarians use as evidence. In the imperfective forms, the helping verb joins the adjective, so Roberts' dzhobl-áw-em "I am injuring" is one word. In the perfective, the two parts stay apart and a clitic can stand between them. Roberts, working from Yusufzai speech, gives dzhóbel krr-em dee and dzhóbel dee krr-em, both "you injured me", with the clitic dee in either place. For a verb in the imperfective, the clitic may not go between the parts. [4] David reports the same freedom for prefixed verbs, where a clitic or the negative can separate the prefix from the stem. [3]

What the tenses do

The present imperfective covers the simple present and the present continuous. Tegey and Robson give gaḍégəm "I dance, I am dancing". [2] David adds that it takes in states and conditions as well as actions in progress or habitual. [3] Tegey and Robson say that in present-time statements the present perfective occurs only in clauses; its main use is the future with bə. [2]

Pashto marks the future with that small word. Tegey and Robson call bə the future marker and say most future expressions pair it with the present perfective. [2] David disagrees with the label. She follows Penzl in treating bə as modal, not as a future clitic, because with a present perfective verb it also covers speculation and doubt, and with a past imperfective it means "would", a habit in the past. She glosses it "WOULD". Present imperfective verbs also express the future, especially when an event is repeated or goes on. [3] Tegey and Robson's sentences:

تور به يې بوزي.

tor bə ye bózi

"Tor will take him."

تور به لاړ شي.

tor bə láṛ si

"Tor will go."

For a negative future, the order of the pieces is fixed. Tegey and Robson give ahmad bə wə nə gaḍegi, "Ahmad will not dance", with bə, then the perfective wə, then the negative nə. [2]

In the past, the imperfective describes events that continued or were repeated, and David gives "all the men were dancing at the picnic, or used to dance at picnics". The perfective "expresses actions with focus on completion, or on the event per se", as in her sentence about Afghans in Bavaria who protested for the release of Afghan prisoners. [3] The perfect tenses, made from the past participle and the verb "to be", are a separate set. Tegey and Robson say the participle agrees with the object if there is one. [2]

The ergative past

In a present-tense sentence the subject is in the direct case and the verb agrees with it. In the past tense of a transitive verb the order of agreement changes. Tegey and Robson list three facts: the subject is in the oblique case, the object is in the direct case, and the verb agrees with the object in person, number and, in the third person, gender. [2] Raverty described the same pattern in 1867: intransitive verbs agree with the nominative, while transitives "are governed, both in gender and number, by the objective case". He called the case of the agent "instrumental". [1]

Tegey and Robson set a present sentence beside its past: [2]

سړی مڼه خوري.

saṛay maṇá khwri

"The man is eating the apple."

سړي مڼه خوړله.

saṛi maṇa khwaṛə́la

"The man was eating the apple."

The man changes from saṛay to saṛi, the oblique form. The verb ending follows the apple, which is feminine, and not the man. In their next pair the scarf is masculine, so the past verb ends differently from the one for the apple. [2] Tegey and Robson add that in the past tense weak object pronouns drop and weak subject pronouns take the oblique form. [2]

مڼه خورم.

maṇa khwrə́m

"I'm eating the apple."

مڼه مې خوړله.

maṇa me khwaṛə́la

"I was eating the apple."

David describes the pattern as nominative-accusative in the non-past tenses and ergative-absolutive in the past tenses. [3] Her tables say a transitive subject is oblique in the past tenses and an intransitive subject direct. [3] Her example of a past aorist, in her own transliteration, is ma der xwandawar tut-an wa-xor-al, "I ate a lot of tasty mulberries". The verb ends in the plural masculine because the mulberries are plural masculine. [3]

Tegey and Robson note that Pashto is an ergative language only in the past tenses. They write that historical linguists suppose all the Iranian languages once had this construction and lost it at different speeds, so that now "only Pashto, Beluchi, and Kurdish retain ergative elements". [2] They add that Pashto appears to be the most conservative of the Iranian languages. [2]

Pronouns add a wrinkle. Roberts quotes Farooq Babrakzai's 1999 dissertation for sentences in which a first or second person pronoun is oblique when it is the object of a present verb: ze taa daftar ta leeg-em, "I am sending you to the office". In the past the roles turn round. minee ze pe baagh kee we lid-em, "Meena saw me in the garden", has "I" in the direct case and a first-person verb. maa mina pe baagh kee we lid-a, "I saw Meena in the garden", has "I" in the oblique and a feminine verb for Meena. [4]

Roberts argues that compound verbs show something more complex than a simple split by tense. In the past perfective of a compound, both the adjective and the helping verb agree with the object, as in sangin kerkey maat-a krr-a, "Sangin broke the window". In a present perfective or a future the adjective agrees with the object while the helping verb agrees with the subject: sangin ba kerkey maat-a krr-i, "Sangin will break the window". He concludes that Pashto keeps an element of ergativity tied to aspect as well as one tied to tense. [4]

Little words that wait

Tegey and Robson call the particles, also named clitics, probably the most famous or infamous aspect of Pashto grammar after the ergative. They are the future marker bə (به), a word for "but, well, then" kho (خو), the weak pronouns me, de, ye, am and mo, and de (دي) in the sense "must". They are placed directly after the first stressed element in a sentence or clause. [2]

The first stressed element can be a subject, an object, a phrase with a preposition, an adverb, a verb or even part of a verb. Tegey and Robson print نه مې پېژني, nə́ me pézani, "He doesn't know me", and و مې پېژانده, wə́ me pezānde, "I recognized him". In the second, the clitic stands between the wə́- and the rest of the verb. [2] David gives the same pattern, from Tegey (1977) and Kopris (2009): wə́ me wah-a, "I beat him". [3] Roberts adds that a clitic needs a host on its left strongly enough that a one-word sentence puts the verb first: wah-i mee, "He hits me". [4]

When more than one particle is present, their order among themselves is fixed. In Tegey and Robson's order, kho comes first, then bə, then me or mo, then de, then ye. [2] David gives the same order, which she attributes to Tegey (1977). [3] The fixed order has a cost, and Tegey and Robson show it in an ambiguous sentence: [2]

پلار مې دي لېږي.

plār me de légi

"My father is sending you." or "Your father is sending me."

They say the ambiguity is seldom a problem in context and that a strong pronoun settles it when it is. Two identical-sounding clitics cannot stand together, so "My father must send you" cannot be built with de twice; the weak "you" gives way to the strong pronoun tə. [2]

Where grammarians differ

The terms are unsettled. Tegey and Robson say in their preface that "names of the tenses are by no means definitely agreed on". [2] In Pashto itself, Mujawir Ahmad Ziar's grammar of 1979 names past tenses with Pashto words, for example پوښتنيز تېرمهال, with the Arabic term in brackets beside it. [5] For the aspects, Penzl and Heston used the Roman numerals I and II, Septfonds the numerals 1 and 2, and David chose "continuous" and "aorist". [3] On bə, Tegey and Robson call it the future marker and David calls it modal. [2, 3]

On the ergative, the grammars agree on the facts and differ on the frame. Raverty called the agent's case instrumental, Tegey and Robson call it oblique, and Roberts argues that compound verbs show aspect as well as tense at work. [1, 2, 4]