This is not a complete list of verbal tenses found in all languages. It focuses on tenses that are expressed synthetically, i.e. using affixes in one word (as opposed to analytical expression, using combination of a main verb and an auxiliary verb). Morphological / part-of-speech tags usually only need to distinguish synthetical verb forms.

• infinitive
• finite verb
  • indicative
    • future
    • present
    • past
      • aorist / preterite
      • imperfect
      • perfect
      • pluperfect
      • preterite anterior
  • imperative
  • conditional
  • subjunctive
  • jussive
• participle
  • present
  • past
  • passive
• transgressive
• gerund
• supine

Usually considered base form of the verb, appears in dictionaries. In text it appears as part of analytical verb forms and as argument of other words:

• future tense: bude dělat, er wird kommen
• argument of modal verb: musí dělat, er muss machen, he must do
• argument of some other verbs, subordinating conjunctions etc.: potřebuje odejít, er kommt um das zu beenden, he let us to do it, his intention to finish this

In some languages infinitive is marked by verb affixes:

• Czech infinitive ends in -t, -ti, -ct, -ci: dělat, dělati, říct, říci
• German infinitive ends in -en: machen, sagen

In some languages, the morpheme marking infinitive is separate word (English to). In English, the base form of the verb is not (full) infinitive. It becomes infinitive when combined with the infinitive mark (to do, to say). Without the mark, it can be used (in English) as imperative.

Both approaches can be combined. Although infinitives are recognizable by the -en suffix in German, the infinitive mark zu is used with infinitives in some contexts. Depending on the verb, zu is either a separate word, or a morpheme inserted between the verb prefix and stem (zu beenden vs. anzupassen).

• English infinitive mark is to
• German infinitive mark is zu
• Danish infinitive mark is at
• Swedish infinitive mark is att

In DZ Interset, infinitive is decoded as verbform = inf. This value is set even for the English base verb form, although it is not certain whether the form is used as infinitive, or imperative. Mood, tense and voice are usually empty. Czech infinitives set the feature negativeness. Portuguese infinitives can be mildly conjugated. I think I may have seen active/passive infinitives but I am not sure where. [in Encyklopedický slovník češtiny p.180? být psán, být bit]

Finite verb forms are those most frequently used. They are classified according to mood: the most ordinary mood is indicative, most languages also have imperative, some also have subjunctive and jussive.

Opposed to finite verbs are non-finite verb forms, namely infinitive, supine, participle, transgressive and gerund.

In DZ Interset, finite verb form is decoded as verbform = fin.

Verbs in indicative mood merely indicate that someone performs / has performed / will perform the action of the verb. Opposed to indicative is imperative (the verb orders the listener to perform the action) and various conditional constructions (the action would be performed if…)

There is a multitude of indicative verb forms, classified according to tense (present, past, future etc.) and voice (active or passive). Present active forms are the default. They are usually the simplest, formed synthetically. Other tenses and voices are formed synthetically or analytically, depending on language.

In DZ Interset, indicative is decoded as mood = ind. This is also the default mood value for finite verb forms. (For non-finite verb forms, default mood is empty.)

Verbs in present tense indicate an action that is taking place at the time of speaking, or in a general timespan (“usually”). They are usually formed synthetically, and take different forms according to person and number, possibly also gender. If two verbs use different morphemes to express the same combination of person, number, gender etc., they are assigned to different conjugation classes. Example: Czech verb nést:

In DZ Interset, present tense is decoded as tense = pres.

Czech verbs are divided lexically to perfective and imperfective. Present tense forms of perfective verbs are in fact forms of future tense. If the physical tagset encodes these forms as present, they will be decoded into Interset as tense = pres. However, if the physical tagset contains the information that these are semantically future forms, it will be decoded as tense = fut.

The English progressive present tense (he is doing) is formed analytically by the simple present form of the auxiliary verb to be and the present participle (gerund) of the main verb. There is thus no Interset feature value for this tense. See participle and gerund.

Simple past indicates action that took place in past. Some languages (e.g. Czech) have only this one past tense. Others have more refined tense system that allow to express whether the action has been completed (perfect) or not (imperfect), whether it was running while something else was being done etc. In Czech however, the perfect/imperfect aspect is lexically inherent, so that past perfect is replaced by simple past tense of a perfective verb.

Czech past tense is formed by past participle of the main verb and (in 1st and 2nd person only) present form of the auxiliary verb být. The participle is decoded with tense = past. (The Czech past tense is a preterite by usage, although its creation bears charactersitics of a perfect, from which it has historically evolved.)

German imperfect past tense (Präteritum) is formed synthetically. Perfect past tense (Perfekt) is formed analytically by present form of the auxiliary verb haben and past participle of the main verb. Pluperfect past tense (Plusquamperfekt) is formed analytically by imperfect past form of the auxiliary verb haben or sein and past participle of the main verb. The imperfect past forms and the participle are decoded with tense = past.

The same holds for English:

In Portuguese, three of five past tenses are synthetic and two are analytic. All three are decoded as tense = past. They have to be further distinguished by the values of subtense. Imperfeito (imperfect) is decoded as subtense = imp. Pretérito (preterite) is decoded as ??? Mais-que-perfeito (pluperfect) is decoded as subtense = pqp. Analytic tenses are perfeito (perfect), tenho caçado, auxiliary in present and past participle, and another form of mais-que-perfeito (pluperfect), tinha caçado, auxiliary in imperfect and past participle.

This is the default past tense. Grammars of some languages call it just “past” or “simple past”. Portuguese calls it pretérito while Spanish calls it pretérito perfecto simple (Czech minulý čas dokonavý), as opposed to pretérito perfecto compuesto, which is parallel to what we call perfect here, see below. Often synthetic. Simply indicates an action that took place in past. No information whether the action has been completed or whether it ran simultaneously with another action.

German example: Präteritum of singen (to sing):

English example: past tense of to sing:

Portuguese example: pretérito of caçar:

DZ Interset decoding: tense = “past”; subtense = “”;

The simple, aspect-neutral past tense is called aorist in some languages, e.g. in Bulgarian. So far I have not come across differences between aorist and preterite, nor have I found a tagset/language where both these tenses would co-exist. The German preterite, for instance, displays usage pattern typical to aorist, although it is not called aorist.

Bulgarian (http://en.wikipedia.org/wiki/Bulgarian_verbs#Past_Aorist_Tense_.28Aoristus.29): Past aorist tense (Минало свършено време) expresses an action that happened at a specific time in the past. Both imperfective and perfective verbs have such tense (there is no difference in their conjugation). Although imperfective and perfective verbs are conjugated in the same way in the past aorist tense, there is difference in their meaning. Compare the sentences: Вчера четох една книга (aorist of imperfective verb) = “Yesterday, I read a book but I didn't finish it”; Вчера прочетох една книга (aorist of perfective verb) = “Yesterday, I finished reading a book/I read a whole book.” The same meaning difference holds for the preterite of Czech imperfective and perfective verbs: Včera jsem četl jednu knihu vs. Včera jsem přečetl jednu knihu.

Example: aorist forms of the Bulgarian verb чете (to read):

DZ Interset decoding: tense = “past”; subtense = “aor”; If we decide that aorist is the same tense as preterite we should handle them both the same way, i.e. tense = “past”; subtense = “”;.

Aorist has almost vanished from Czech. The only survivors are aorist forms of the auxiliary verb být (to be):

Since these forms are today used exclusively to construct analytic forms of conditional, they are decoded as conditional / subjunctive: mood = “sub”.

Czech term: souminulý čas. Indicates that the action was taking place at the same time as some other action in past, similarly to the English construction I was doing. The English construction is analytic, made of past (preterite) of the auxiliary to be, and of present participle (gerund) of the main verb. However, there are languages with synthetic imperfect forms.

Portuguese example: imperfeito of cantar (to sing):

Spanish example: imperfecto of cantar (to sing):

Bulgarian (http://en.wikipedia.org/wiki/Bulgarian_verbs#Past_Imperfect_Tense_.28Imperfectum.29): Past Imperfect Tense (Минало несвършено време) is used to talk about a temporary situation that existed at or around a particular time in the past. It also expresses past actions that were frequent, repeated, permanent or always true. Its most common use is in story telling to provide a background to other actions which are usually expressed with verbs in the past aorist. In this use it means that the action had begun and was in progress when the other action(s) happened, we don't know whether it stopped or not.

Both imperfective and perfective verbs have past imperfect tense. They are conjugated in the same way.

Verbs form the past imperfect tense with the following endings. Example: чете (to read):

DZ Interset decoding: tense = “past”; subtense = “imp”;

Czech term: předpřítomný čas. A tense (not to be confused with perfective lexical aspect in some languages) indicating an action that took place in past but its relation to the present time point (result) is stressed (typically that the action has already / has not yet been completed). It is also called present perfect, as opposed to past perfect (pluperfect) and future perfect. In English, the exact time when the action took place is unknown, so e.g. the word yesterday cannot occur in a clause headed by perfect verb form.

Perfect is usually created analytically and thus there is no feature value denoting it. Typical formation is using present tense of an auxiliary and past participle (perfect participle) of the main verb.

DZ Interset decoding: auxiliary: tense = “pres”; participle: verbform = “part”; (tense = “past”)

Example: present perfect of the English verb to do:

Example: present perfect of the German verb machen (to do):

Example: present perfect (perfeito) of the Portuguese verb cantar (to sing) is formed by present indicative of the auxiliary verb ter and the past participle (participio) of the main verb:

Example: present perfect (минало неопределено време, past indefinite tense) of the Bulgarian verb чета (to read) is formed by present indicative of the auxiliary verb съм (to be) and by past active aorist participle of the main verb:

Czech term: předminulý čas. A tense indicating an action that took place in past and has been completed before another action in past. It is also called past perfect, as opposed to (present) perfect and future perfect.

Pluperfect is usually created analytically; however, see Portuguese synthetic pluperfect below. Typical formation is using past (preterite) tense of an auxiliary and past participle (perfect participle) of the main verb.

Example: past perfect of the English verb to do:

Example: past perfect (Plusquamperfekt) of the German verb machen (to do):

Portuguese pluperfect (mais-que-perfeito) has both synthetic and analytic forms. The analytic forms are constructed similarly to Germanic languages, from past (imperfeito) tense of the auxiliary verb ter and the past participle (participio) of the main verb. Note that the synthetic forms are parallel to Spanish forms that are called subjunctivo imperfecto (spojovací způsob souminulý). Portuguese also knows subjuntivo imperfeito but its forms are neither homonymous, nor similar to the synthetic pluperfect. Spanish has the analytic pluperfect forms (pretérito pluscuamperfecto).

Example: past perfect (минало предварително време, past preliminary tense) of the Bulgarian verb чета (to read) is formed by past aorist indicative of the auxiliary verb съм (to be) and by past active aorist participle of the main verb:

DZ Interset decoding of the analytic forms: auxiliary: tense = “past”; participle: verbform = “part”; (tense = “past”). Decoding of the synthetic forms in Portuguese: tense = “past”; subtense = “pqp”;.

Spanish: pretérito anterior. Spanish bookish pluperfect forms that are used in tandem with the simple perfect forms (while normal pluperfect is used to compare time with imperfect past action). In colloquial language only normal pluperfect is used. Pretérito anterior (Czech předminulý čas dokonavý) is formed analytically by simple perfect forms of the auxiliary verb haber and the participle of the main verb: hube hablado, hubiste hablado…

Participle in compound verb tenses has only one form (ella ha cantado), in contrast to adjectival use where gender and number is distinguished (casado hombre, casada mujer, casados hombres, casadas mujeres). Gerund (gerundio) has only one form (ella está cantando).

Grammars of some languages regard conditional as a tense, although others treat it as a mood. Czech: podmiňovací způsob (conditional mood). In English and other languages the present conditional can indicate future relative to some point in the past: I warned him that I would call the Police if he did not turn down the music. Often constructed analytically, in which case there is no need to use mood = “cond”.

Czech: present conditional mood is constructed by aorist form of the auxiliary verb být (to be) and the active past participle of the main verb. Past conditional mood is constructed from the present conditional of the auxiliary verb být and the active past participle of the main verb.

English: conditional is made of would (past tense of the auxiliary verb will) and of base form of the main verb. Past conditional is made similarly but instead of the present indicative, there is the compound present perfect form of the main verb. Example: verb to sing.

German: conditional (Konditional) is made of past subjunctive (Konjunktiv Präteritum) form of the auxiliary verb werden, and the infinitive of the main verb.

Portuguese: present conditional (condicional) has synthetic forms and past (perfect) conditional is analytic, made of present conditional of the auxiliary verb ter, and of the participle of the main verb.

The synthetic present forms require that the decoder set mood = “cond”.

In grammars of some languages called conjunctive. Czech: spojovací způsob. In some languages its usage overlaps with conditional (German Konjunktiv: Hätte er heute kommen wollen, würde ich nicht abreisen. = “If he wanted to come today I would not travel away.” Spanish subjuntivo is used in subordinated clauses after a subordinating conjunction like in que hable = that I speak = (Czech) ať mluvím, abych mluvil. It also works as a sort of imperative in persons that do not have their own imperative forms (1st and 3rd).

German has synthetic forms called Konjunktiv I (present) and Konjunktiv II (past, preterite). There is a significant overlap of forms of present subjunctive with present indicative.

Portuguese forms subjunctive (subjuntivo) of 6 tenses, 3 of that synthetic.

DZ Interset decoding: mood = “sub”

Unordered observations from the Ancient Greek Dependency Treebank:

present:
v v pos=v|per=3|num=s|ten=p|mod=i|voi=a|gen=-|cas=-|deg=- (ἔχει, ἐστὶ, ἐστιν, χρὴ, ἔστι)

imperfect:
v v pos=v|per=3|num=s|ten=i|mod=i|voi=a|gen=-|cas=-|deg=- (προσέφη, προσηύδα, ἦν, ἦεν, ηὔδα)

perfect:
v v pos=v|per=3|num=s|ten=r|mod=i|voi=a|gen=-|cas=-|deg=- (ἔοικε, ἔοικεν, οἶδεν, οἶδε, οἶδ’)

pluperfect:
v v pos=v|per=3|num=s|ten=l|mod=i|voi=a|gen=-|cas=-|deg=- (ὀρώρει, βεβήκει, ᾔδη, ἑστήκει, ἐῴκει)

future perfect:
v v pos=v|per=2|num=s|ten=t|mod=i|voi=a|gen=-|cas=-|deg=- (ἀγορεύσεις)

future:
v v pos=v|per=3|num=s|ten=f|mod=i|voi=a|gen=-|cas=-|deg=- (δώσει, ἕξει, μελήσει, ἐρέει, ἐρεῖ)

aorist:
v v pos=v|per=3|num=s|ten=a|mod=i|voi=a|gen=-|cas=-|deg=- (βῆ, προσέειπε, ἦλθε, βάλε, στῆ)

Příklad: sloveso ἀγορεύω:
v3siia = ἀγόρευε, ἀγόρευεν, ἀγόρευ’
v3saia = ἀγόρευσε, ἀγόρευσεν
v3spia = ἀγορεύει
v3spoa = ἀγορεύοι
v2stia = ἀγορεύσεις