9. Verbs

     *** THIS PAGE IS BEING REVISED ***

Notes:

1. I have departed from IPA in the representation of mid vowels, and follow the long established academic practice when describing Vulgar Latin and the development of Romance languages, namely:

   [e]and [o] denote any front unrounded or back rounded mid vowel respectively, when the degree of height is either unknown, indeterminate or irrelevant;

   [ẹ] and [ọ] denote specifically high versions (IPA [e] and [o]);

   [ę] and [ǫ] denote specifically low versions (IPA [ɛ] and [ɔ]).

2. Also [ə] denotes:

   - generally any reduced non-defined centralized vowel without reference to height;

   - in Late Britainese, a unstressed non-phonemic low-mid to near-low ([ɜ] to [ɐ]) central vowel, depending upon regional pronunciation.

3. In discussing the finite verb forms below the following abbreviations are used:

   1SG = 1st person singular 2SG = 2nd person singular     3SG = 3rd person singular

   1PL = 1st person plural 2PL = 2nd person plural 3SG = 3rd person plural

9.1 Conjugations in Vulgar Latin

Verbs are usually the most complicated part of speech, especially so in the Romance languages. Those who know Latin, will recall there were four conjugations in Classical Latin, thus:

• 1st amō, amāre "to love"
• 2nd videō, vidēre "to see"
• 3rd bibō, bibere "to drink"
• 4th dormiō, dormīre "sleep"

and the 3rd conjugation having a subset of facio, facere "to make, to do."

The latter small subset in fact have a mix of 3rd and 4th conjugation endings. They hint at something that was going on spoken Latin, namely some confusion of conjugations. This shows up also in the Classical language in verbs which some authors assign to the 2nd conjugation and others to the 3rd, e.g. ferveō, fervēre or fervō, fervere "to boil, ferment, glow"; tergeō, tergēre or tergō, tergere "to wipe, clean.".

In fact it is clear there was considerable fluctuation and confusion in the spoken language between the 2nd, 3rd and 4th conjugations, probably helped by /e/ and /i/ often being pronounced similarly in unstressed post-tonic syllables. The first conjugation, however, generally held its own.

Of the anomalous verbs:

• volō, velle "to wish, want" joined the 2nd conjugation: *voleō, volēre;
  (its two Calssical derivatives, mālo, mālle and nōlō, nōlle did not survive in Vulgar Latin.)
• possum, posse "to be able" was partially assimilated to the 2nd conjugation, but retained some anomalous forms thus: possō, potēre;
• ferō, ferre did not survive, giving way to portō, portāre "to carry";
• fīō, fierī "to become, happen" was regularized to 4th conjugation but survived only in Balkan Romance where it is suppletive with descendants of Latin esse;
• Some forms of eō, īre "to go" survived in Vulgar Latin but with extensive suppletion with other verbs, especially vādō, vādere "to go";
• sum, esse "to be" acquired a 3rd conjugation infinitive ending essere [ˈɛs.sɛ.ɾɛ] but otherwise remained anomalous.

The entire passive inflexions were dropped in Vulgar Latin, being replace partly by reflexive constructions and partly by esse(re) with the perfect passive participle. Deponent verbs simply became normal active verbs, e.g. mentīrī → mentīre, whence French, Occitan, Catalan, Spanish, Portuguese and Britainese mentir "to lie, say something untrue."

9.2 Forms surviving in the Proto-Romance of Gaul and Britain

9.2.1 Finite Verbs

Finite verbs function as the head of verbal clauses and show time (tense), aspect and mood; in Latin, the Romance languages and many others they are also marked by person and number. In Latin, the Romance languages and many others time and aspect are not clearly distinguished, e.g. the Latin "Perfect Tense" may show either past time and perfective aspect or present time and perfect aspect according to context. In strict linguistic terminology tense denotes time difference; but below and in the rest of this page I use Tense (with upper-case T) in the sense used in traditional grammar to denote time, aspect and mood paradignms found in Latin and the Romance languages.

9.2.1.1 Indicative Tenses

The Present, Imperfect, Perfect and Pluperfect survived. The Perfect, however, lost its present perfect meaning, this being expressed by a compound of habēre and the perfect participle if the verb was transitive or esse and the perfect participle if intransitive, e.g. fēcī → factum habeō "I have done (it)"; vēnī → *venūtus sum "I have come." The Latin Perfect tense retained only its past perfective use, i.e. it became the Preterite tense. The Latin Pluperfect was retained as a Preterite or as a Conditional in various regions but dropped out of use at an early date.

Only esse retained its future erō, eris ..; with other verbs futurity was expressed by suffixing contracted forms of habēre to the infinitive, e.g. cantāre + *a(b)jō. Eventually even esse(re) developed similar forms. Parallel with this a new "Future in the Past" or "Conditional" was developed by suffixing contracted forms of the Imperfect of habēre to the infinitive.

The old Future Perfect was confused with the Perfect Subjective and did not survive in Gaul or Britain

9.2.1.2 Subjunctive Tenses

Of the four Latin subjunctive tenses, only the Present and Pluperfect survived in Gaul, the latter simply becoming simply the Past Subjunctive; there is no reason to suppose that this would not have also been the case in Britain.

9.2.1.3 Imperatives

The Latin "Future Imperatives" with their 2nd and 3rd person forms did not survive in Vulgar Latin. Only the 2nd persons of the Present Imperative survived, and the plural was often expressed by the 2nd person plural of the Present Indicative tense. The 1st person hortative and 3rd person jussive meanings were expressed by the Present Subjective.

9.2.2 Nonfinite Verbs

Nonfinite verbs cannot function as the head of an independent clause; they combine verbal functions with that of nouns, adjectives or adverbs.

9.2.2.1 Supine

The Latin supine was the accusative, e.g. spectatum veniunt "they are coming to watch", and ablative (or possibly dative), e.g. difficile dictū est "it is difficult to say", of an old 4th declension verbal noun. It did not survive in Vulgar Latin.
(The supine of modern Romanian has a superficial resemblance to the Latin supine, but it not attested before he 16th century, and then only rarely until the 17th century. It is a relatively modern development and originated from nominalized past participle forms of intransitive verbs as Professor Virginia Hill explains in her paper "The emergence of the Romanian supine".)

9.2.2.2 Gerundive

The gerundive was a passive verbal adjective showing necessity or obligation, e.g. epistola scrībenda est "the letter is to be written (i.e. the letter must be written)". It did not survive in Vulgar Latin.

9.2.2.3 Infinitive

The infinitives in Latin were verbal nouns that could function as the subject or direct object of a sentence or as a complement after "to be", e.g. labōrāre nōn amat "he doesn't like to work/ he doesn't like working", labōrāre est ōrāre "to work is to pray/ working is praying." It could not be used after a preposition, nor in any case other than the nominative and accusative. If one wished to have a verbal noun in any other case or to use it after a preposition, one had to use the Latin gerund (see below).

In Classical Latin verbs could have three synthetic infinitives:

• present active, e.g. amāre "to love";
• present passive, e.g. amārī "to be loved";
• perfect active, e.g. amāvisse/ amāsse "to have loved".

Text books give other infinitives formed from compounds; these do not concern us here.

Of the three synthetic infinitives, only the present active survived in Vulgar Latin; it survives also in the modern Romance languages where its use has greatly expanded and it is commonly used after prepositions, cf. French le plaisir de voyager "the pleasure of travelling"; facile à dire "easy to say."

9.2.2.4 Gerund

The Latin gerund was an active verbal noun that, so to speak, filled the spaces the present infinitive could not, i.e. it was used in the accusative and ablative cases after prepositions, and in the genitive, dative and plain ablative. In theory, with the collapse of the case system and the use of prepositions before the present infinitive, one would have expected the gerund to have become defunct and not to have survived into Proto-Romance.

The ablative of the gerund came in the writers of the Empire to function as a present participle, e.g. novī cōnsulēs populandō usque ad moenia pervēnērunt (Livy 8.17.1) "the new consuls, laying waste [the land], came right up to the city walls"; exturbābant agrīs, captīvōs servōs appellando (Tac. _Ann._ 14.31.2) "they ejected [people] from their farms, calling them captives [and] slaves." This usage gained ground in Vulgar Latin, cf. in Egeria's (or Etheria's) 4th centuray account of her pilgramage to the Holy Land we find redīre ... dīcendō psalmōs "to return ... saying psalms" and remains so to the present day in Italian, Spanish and Portuguese.

In Catalan and the Romance languages of Gaul, the ablative lost its final vowel and the final /d/ become unvoiced in word final position, giving gerunds ending in -ant and -ent.

9.2.2.5 Participles

Participles are verbal adjectives. In Classical Latin a verb may have three such participles:

1. Future active participle perhaps best known in the greeting of gladiators to the Emperor Claudius according to the writer Suetonius: Havē Imperātor, moritūrī tē salūtant "Hail Emperor, those about to die salute you." It is rare in Vulgar Latin and did not survive in Proto-Romance.
2. Present active participle was inherited from Proto-Indo-European but in Old Latin had already lost most of its verbal functions and was mainly used as an adjective, e.g. vigilāns "awake, alert", maerēns "sorrowful", sedēns "seated", sapiēns "wise, discerning." This remained true of colloquial Latin and those Latin present participles which survive in the Romance languages survive as adjectives, not as participles.
   (In the written language of the Classical period of the late Republic, the participle did become used as a true verbal adjective; this was undoubtedly due to Greek influence as ancient Greek but this did not become part of the colloquial language where, as we have seen above, the ablative of the gerund filled the function of the present active participle.)
3. Perfect participle which, in Classical Latin, was normally passive, the main exceptions being the perfect participles of deponent verbs, e.g. secūtus "having followed" (← sequī "to follow") and a few other verbs, e.g. cēnātus "having dined" (← cēnāre "to dine"). Classical Latin intransitive verbs did not normally have a perfect participle except in impersonal passive constructions, e.g. postquam perventum est "after they/people arrived".
   The perfect participle was retained in Vulgar Latin and greatly increased in that intransitive verbs acquired participles with active meaning so, e.g. the old impersonal ventum did not survive as a perfect participle of venīre "to come", but was replaced by *venūtus "having come" in Gaul and Italy or •venītus "having come" in the Iberian peninsula and Dacia.

The only participle that survived in Vulgar Latin and Proto-Romance was the perfect participle (active in the case intransitive verbs and passive if the verb was transitive). Indeed, it not only survived but extended its use in two three ways:

• In Classical Latin the perfect passive participle was used with the verb "to be" to form the Perfect, Pluperfect and Future Perfect Passive (e.g. amāta est "she was loved, has been loved", amāta erat "she had been loved", amāta erit "she will have been loved"); but in the spoken language these shifted their meaning with the participle simply being a passive participle, i.e. amāta est "she is loved", amāta erat "she was loved", amāta erit "she will be loved."
• In Classical Latin the Perfect, Pluperfect and Future Perfect of deponent verbs had been formed by using the perfect active participle with "to be"; deponents, as we have seen, disappeared from the spoken language but this method of forming the perfect active tenses was extended to all intransitive verbs, e.g. *venūta/venīta est "she has come", *venūta/venīta erat "she had come", *venūta/venīta erit "she will have come."
• The latter was not possible with transitive verbs as the perfect participle was passive. However we find as early as Plautus (c. 254 – 184 BC) multa bona bene parta habēmus "we have many well procured goods" which could be understood as "we have procured many goods well"; this construction. habēre and perfect passive participle agreeing with the direct object, developed in spoken Latin a periphrastic way of expressing the Perfect, Pluperfect and Future Perfect tenses of transitive verbs. Thus we find, for example, in Greogory of Tours episcopum invītātum habēs "you have invited the bishop."

9.2.3 Summary

Proto-Britainese will have inherited these forms:

• Vulgar Latin Present, Imperfect and Perfect Active Indicative tenses;
• Future and Conditional (Future in the Past) Active Indicative tenses formed from suffixing contracted forms of habēre to the Present Active Infinitive:
• Present and Past Active Subjunctive tenses derived from the Latin Present and Pluperfect Subjunctives respectively;
• 2nd Person Singular Active Imperative;
• Present Active Infinitive;
• An invariable Gerund ending in -ndo;
• Perfect Participle, Active in the case of intransitive verbs, and Passive if verb is transitive.

9.3 The Present Indicative Tense

Those familiar with Romance languages will know that if verbs are irregular that it is more often than not the Present Indicative that is irregular; thus we begin with this tense. I must stress, however, that not all irregular verbs are given; subsections 9.3.3 and 9.3.4 just show examples of the sort of irregularities that may occur. Subsection 9.3.5 gives auxiliary verbs (and they can be guaranteed to be irregular!).

Also it must be remembered that while, e.g. French il chant can mean either "he sings" or "he is singing", in Britainese il cant can mean only "he sings" and the meaning "he is singing" is expressed by a periphrastic construction similar to English and Welsh.

9.3.1 Regular verbs

In Vulgar Latin the 1SG endings of the 1st and most 3rd conjugation verbs was -ō; the 2nd conjugation -eō and 4th conjugation -iō (and a few 3rd conjugation verbs) would be expected to become [jo] in Vulgar Latin but generally simply became [o] in line with the 1st and 3rd conjugation verbs.

The 4th conjugation 3PL -iunt (found also with a few 3rd conjugation verbs) simply became [ont] in Vulgar Latin.

All the singular endings and the 3PL ending were unstressed in Latin and this remained so in Vulgar Latin and all the Romance languages. The 1PL and 2PL, however, were stressed in in the 1st, 2nd and 4th conjugations. It seems that in Proto-Romance there was a shift to stressing those endings in the 3rd conjugation also. Therefore we may expect the following in early Britainese:

	1st conj.	2nd conj.	3rd conj.	4th conj.
1SG	-
2SG	-es [əz]	-s [z]
3SG	-ed [əd]	-d [d]
1PL	-ams ['amz]	-eims ['ei̯mz]		-ims ['imz]
2PL	-ads ['adz]	-eids ['ei̯dz]		-ids [idz]
3PL	-ent [ənt]	-nt [n̩t]

9.3.1.1 The 1SG, 2SG & 3SG

All regular verbs had no ending for the 1SG and when [ə] fell silent in final syllables, all regular verbs behaved the same way.

The 2SG ending -s was normally pronounced [z], but was [s] after voiceless consonants, e.g. mets ← metre "to put, place." If the verb stem already ended in -s another -s was not added, e.g. eu conoiss ← *connokso ← cognōscō "I know (of)", tu conoiss ← *connoksis ← cognōscis "you [s] know (of)." Also the ending -s is not separately pronounced when suffixed to verbs ending -sce [ʃ], -ce [t͡ʃ], -ge [d͡ʒ]. -sie [ʒ], -x [ks] and -z [z], e.g. eu basie [əu̯'baʒ] ~ tu basies [ti'baʒ] "I kiss ~ you [s.] kiss" and, indeed, the final -s was not always written. Indeed, Forms such as conoiss and 'inchoative' verbs like eu flouriss, tu flouriss "I prosper, you prosper" led to final -s of the 2nd person being more and more dropped as the compulsory subject pronoun sufficiently marked second person singular and, unlike the plural -s ending on nouns and pronouns, the grammatically redundant.

In early Britain, 3SG final [d] would have been devoiced after voiceless consonants and written as -t, e.g. el conoist [əlku'nwist] "she knows." After dental/alveolar plosives, the final -d would have been fully assimilated e.g. vend+d → vend "sells", part+d → part "departs, leaves." What happened after [ð] would have been unclear in old texts due to the various ways [ð] was spelled (or not spelled); although we have the combination [ðd] in words like "smoothed" and "breathed", it is made easier in English in that while [ð] is a dental fricative [d] is an alveolar plosive, but we cannot be sure of the precise value of [d] in Britainese.

Also it should be borne in mind that while the suffix ed has grammatical significance in English, i.e. marking the preterite tense and the perfect participle, it had no such function in Britainese which was a non-pro-drop language; it is probable that the final -d was dropped after voiced fricatives e.g. il veidhd → il veidh "he sees" It is also likely that final -d tended also to be dropped after other "difficult" combinations, e.g. beivd → beiv "drinks", dormd → dorm "sleeps." amed → amd "loves" Thus there was during Middle Britainese and early Late Britainese a growing tendency for the 3rd person singular not to be mark, though we might expect final -t in, e.g. conoist to be more resistant. Also recall that though final -t in the 3SG of French verbs was once pronounced but fell silent several centuries ago. Would Britainese have been the only Romance language besides Sardinian to have preserved this 3SG ending?

It is likely that there would have been confusion and hesitation in attempts to preserve these redundant endings but, like spoken French for the most part, the endings would have been dispensed with and, unlike French, Britainese would not have retained them in the written language.

9.3.1.2 The 1PL

According to F.G. Mohl, La première personne du pluriel en gallo-roman (1900), forms like *mittomus, due to Celtic influence, were used in northern Gaul instead of míttimus, then the accent was shifted to the penultimate *mittómus, whence the Old French endings -omes, -oms, -ons (metomes, metoms, metons) → modern French -ons (mettons). This theory has not found general acceptance. Indeed, I find no support for a Celtic influence generalizing a 1st person plural ending in -omos.

The common view is that the French endings developed from a generalization of the 1PL of "to be" somes, soms, sons ← sumus [ˈsọmọs]. But why that should be so and why it became the 1PL ending for all verbs at a very early date is not at all clear and I see no reason to suppose that Britainese would have behaved the same way.

It will, however, be seen from French and that in this area there was a strong move towards having common forms for the stressed endings of the 1PL and 2PL. For reasons explained in the next subsection, it is likely by the end of the Old Britainese period or in the early Middle Britainese period, -ams had becomes the common 1st person plural ending.

9.3.1.3 The 2PL

Although the early use of -oms, -ons is obscure, the derivation French 2PL suffix -ez is clear; in Old French it was pronounced [etz] and is derived from Latin first conjugation suffix -ātis. In the earliest Old French texts it competed with -eiz [ei̯ts] or -oiz [oi̯ts] of the 2nd & 3rd conjugations, and -iz [its] of the 4th conjugation. Before the end of the Old French period -ez had ousted its rivals. This was probably because verbs of 1st conjugations were more numerous than others and it was also reinforced by the common 2PL ending of the Imperfect which in Classical Latin always ended in -bātis (except "to be" which had erātis). It is reasonable to suppose that something similar happened in Old to Middle Britainese and that the stressed endings -ams, -ads became common to all verbs.

It is possible that in Old Britainese spellings of the 2PL such as -az, -eiz, -iz occurred where z = [dz] as well as -ads, -eids, -ids; certainly contact with French is likely to have encouraged such spellings. In Middle Britainese we can expect -az or -adz besides -ads; it is likely that -adz would have prevailed to distinguish 2PL from the plural of the perfect participle of 1st conjugation verbs. None of these considerations apply to the 1st person plural ending and there is no reason to suppose that it did not remain -ams.

9.3.1.4 The 3PL

We saw in subsection 3.1.2.7 that the Latin 3rd personal plural -nt would have survived in early Britainese and, in our timeline, survives till the present day in literary Welsh and, slightly disguised, in spoken Welsh where the post-clitic subject pronoun has modified things, thus -nt hwy [nthuːɨ̯] → -n nhw [nhuː]. But would it have survived in Britainese, or would it have been dropped as it was in French? Consider:

• Although Welsh has preserved personal endings and both Cornish and Breton preserve them in certain syntactic situations, the latter two language use a single unchanging form in normal affirmative statements thus, e.g.

  BRETON	CORNISH	ENGLISH
  me a wel te a wel eñ a wel hi a wel ni a wel c'hwi a wel int a wel	my a wel ty a wel ev a wel hi a wel ni a wel hwi a wel i a wel	I see you [sg] see he sees she sees we see you [pl] see they see

• It will be seen in the table above that standard English adds no endings except for the 3SG. I recall when in early twenties I was on holiday in East Anglia I heard people giving no ending to the 3SG. Also in Sussex where I grew up the final -s was added to all persons of the present simple tense, i.e. I sees, you sees, we sees, they sees. I had naïvely assumed this was a Sussex peculiarity; therefore when I moved to Newport in South Wales in 1968 I was surprised to find exactly the same there in colloquial speech. I have since learnt that affixing -s to the whole present simple tense is not uncommon in colloquial dialects of southern English. In other words, there is a clear tendency towards a single form for the whole tense.
• In modern French personal endings have to a large extent a purely orthographic value, i.e. they are silent in the spoken language (apart from where they've been preserved or restored in liaison), except for 1PL [õ] and 2PL [ẹ[]

It seems to there is a clear regional tendency towards a single form; therefore it seems to me that Britainese would eventually have lost all the unstressed personal endings so final unstressed -nt would have gone the same way as the singular unstressed endings.

9.3.1.5 Examples of regular present tenses in late Britainese

cantar "to sing"	conoistr "to know (of), be acquainted with"	vendr "to sell"	dormir "to sleep"
eu cant tu cant el/il cant nos cantams vos cantadz els/ils cant	eu conoiss tu conoiss el/il conoiss nos conoissams vos conoissadz els/ils conoiss	eu vend tu vend el/il vend nos vendams vos vendadz els/ils vend	eu dorm tu dorm el/il dorm nos dormams vos dormadz els/ils dorm

9.3.2 -ir verbs with inchoative suffix

Inchoative verbs, sometimes called an "inceptive" verbs, show a process of beginning or becoming; in Latin they were shown with the suffix -sc- which lengthened the vowel before it. We have come across one such verb above, i.e. co(g)nōscō "to get to know, to become acquainted [with]" → [eu] conoiss.

But what we are concerned with here is a phenomenon we find in French, Catalan and Italian where some Latin 4th conjugation verbs behave like verbs derived from the 2nd and 3rd conjugation, but the majority are derived from Vulgar Latin forms with the inchoative suffix -īsc- or -ēsc- in the present tense. It should be noted that the suffix lost its strict inchoative or inceptive meaning; nevertheless, such verbs are still generally referred to as 'inchoative verbs' in Romance studies and that is how it is to be understood here. Cf. the table below.

FRENCH			CATALAN			ITALIAN
dormir	fleurir		dormir	florir		dormire	fiorire
je dors tu dors il/elle dort nous dormons vous dormez ils/elles dorment	je fleuris tu fleuris il/elle fleurit nous fleurissons vous fleurissez ils/elles fleurissent		dormo dorms dorm dormim dormiu dormen	floreixo floreixes floreix florim floriu floreixen		dormo dormi dorme dormiamo dormite dormono	fiorisco fiorisci fiorisce fioriamo fiorite fioriscono

Such inchoative verbs are found in Occitan also; but there the inchoative suffix has spread to verbs which did not originally have it. Thus in Occitan the present tense of dormir is:
dormissi, dormisses, dormís, dormissèm, dormissètz, dormisson.

This type of conjugation seems to have begun in Vulgar Latin with verbs such as florére "to blossom, flower; to flourish, prosper" which by using the inchoative form florésco besides the normal flóreo made it possible to have a present tense with stress on the ending throughout, i.e.
florésco, floréscẹs, floréscẹt, florémọs, florétẹs, floréscọnt.

It seems these verbs tended to be assimilated to the -íre group (i.e. Latin 4th conjugation); it will be seen, indeed, that Romance language descendants of Latin flōrēre have almost all joined the -ir(e) group. We see above Italian keeping close to the Vulgar Latin original and Catalan also preserving the original pattern of keeping stress on the endings throughout. In French we see that at an early date the endings -om(es)s/-ons, -ez came to be used for all verbs; when these endings were extended to 4th conjugation inchoative verbs, these verbs were treated just like other inchoatives such as [eu] conoiss, i.e. the suffix -iss- was retained throughout as we see in the table above; Occitan behaved in a similar way.

There can be little doubt that Britainese would have also had such inchoative formations in the -ir group and, as the endings -ams, -ads/-adz where extended to all verbs, that they would have behaved much as in French, and that we could add to the table above:

BRITAINESE
dormir	flourir
eu dorm tu dorm el/il dorm nos dormams vos dormadz els/ils dorm	eu flouriss tu flouriss el/il flouriss nos flourissams vos flourissadz els/ils flouriss

9.3.3 Change of stem vowel in 1PL & 2PL

In accordance with phonetic laws in Old French, the vowel of the verb stem often developed differently according to whether the stress fell on the stem (1SG, 2SG, 3SG and 3PL) or on the ending (1PL and 2PL). This was certainly true of Old French, e.g.

• amās → aimes ['ɛ.məs] "thou lovest, you (sg.) love"
• amātis → amez [a'mets] "you (pl.) love"

where [ɛ] of the 2nd singular contrasts with [a] of the 2nd plural.

This oscillation between two forms of the stem is known as apophony and was common in Old French. Other examples of change of vowel are: leves ~ lavez, trueves ~ trouvez, espeires ~ esperez, lieves ~ levez, pries ~ preiez, pleures ~ plourez, suefres ~ soufrez. In the majority of cases, the stem of the 1PL and 2PL has been generalized (tu laves ~ vous lavez, tu trouves ~ vous trouvez, tu espères ~ vous espérez, tu lèves ~ vous levez, tu souffres ~ vous souffrez), but sometimes the reverse has happened (tu aimes ~ vous aimez, tu pries ~ vous priez, tu pleures ~ vous pleurez). But quite a few modern French verbs have preserved the earlier apophony (tu bois ~ vous bouvez, tu veux ~ vous voulez, tu dois ~ vous devez, tu peux ~ vous pouvez).

To some extent this is lessened in Britainese in that, as we have seen, vowels are often pronounced differently if stressed or unstressed; cf. eu dorm [əu̯'dorm] ~ nos dormams nuz.dur'manz]. Nevertheless, there is likely to have been some examples of apophony in Old Britainese, such as:

vedhair "to see"	baivr "to drink"
eu vaidh tu vaidh el/il vaidh nos vedhams vos vedhadz els/ils vaidh	eu baiv tu baiv el/il baiv nos bevams vos bevadz els/ils baiv

9.3.4 Auxiliary verbs

In proto-Romance there were three auxiliary verbs:

• éssere "to be" which was used with Perfect Participle of transitive verbs to form passive tenses (an extension of Classical Latin usage) and with the Perfect Participle of intransitive verbs to form the perfect active tenses (a usage found in Classical Latin with deponent verbs);
• stáre "to be [in a certain place, in a certain condition]" which could be used with the ablative of the gerund to express a continuous or progressive meaning, e.g. stat cantando "s/he is singing", stabant currendo "they were running"; but such forms were far less used than the corresponding English forms.
• habére "to have" which was used with the Perfect Participle in agreement with the direct object to form perfect active tenses of transitive verbs.

Several of the Romance languages have retained the two distinct forms of "to be", derived from éssere and stáre respectively; this is especially so the Iberian Romance languages with the notable exception of Arogonese which has only one verb "to be", being a mix of forms derived from éssere and stare. More importantly for us we find a similar thing in French and other langues d'oïl where "to be" is a fusion of forms derived from éssere (Old French estre) and from stáre (Old French ester). The evidence of French, English and Welsh (all of which have a single verb "to be" with suppletive forms) in our timeline strongly suggests the two verbs will have fused as they have in French and other langues d'oïl and Aragonese.

Some Romance languages have used other verbs as auxiliaries, e.g. Catalan uses a variant of the present tense of anar "to go" with the infinitive to express the simple past. Britainese, as we have seen, developed continuous tenses similar English, e.g. "I am singing, you were singing" etc. This is common both to English and to Insular Celtic in our timeline; it would seem to be an areal feature. Both English and the Brittonic languages also use the verb "to do" as an auxiliary which suggests Britainese would likewise have used "to do" (Latin facere) as an auxiliary. Britainese, therefore, has three auxiliary verbs:

• star ← stáre [being suppletion of éssere and stáre - see above]
• avair ← habére
• fair ← */fakʲrə/ ← fácere

9.3.4.1 TO BE

This verb is anomalous in all the Romance languages, in English and in Welsh in our timeline and, indeed, in most European languages. It was anomalous in Latin where its present tense was: sum, es, est, sumus, estis, sunt.

The table below shows the verb in standard French and in langues d'oïl dialects of northern France from Gallo in west to Walloon in the east (the Norman shows Rouen dialect).

FRENCH	GALLO	NORMAN	PICARD	WALLOON
être	estr	ète	esse	èsser
je suis tu es il/elle est nous sommes vous êtes ils/elles sont	je soe/sei t'es il/al est je som/j'estom vóz estt/estaez il/al sont	je sieux/sis t'es il/ol est je/nos sommes vos êtes i/o sont	éj su t'es il/al est (n)os sons os ètes i sont	dji so t'es il/ele est nos estans vos estoz i/ele sont

Almost everywhere 1SG and 3PL of the Romance languages are from Vulgar Latin *sọn (CL sum) and *sọnt (CL sunt); rare exceptions are the remodelling of 3PL, e.g. Romansh èn). Hence, for example, Picard su, sont and Walloon so, sont. In Gaul and the Iberian peninsula there are also 1SG forms derived from *sọjo (remodelled by analogy with *ajo ← *aβjo, see 9.3.4.2 below), hence Gallo soe or sei, also French suis and Norman sieux or sis. The French and Norman 1SG also show a paragogic final -s which started to be added to the 1SG of all except -er verbs from the 12th century before later falling sulent though continuing to be written; this development does not concern us. The vacillation between forms derived from *sọn and *sọjo in the Iberian peninsula with Portuguese sou and Spanish soy. In Old Britainese we can expect the 3PL to have been sount [sunt] and the 1SG sou [su] perhaps with variant s(o)ui [sui̯]; we would expect the latter to become [swi] in later Britainese. But when final -nt of the 3PL was falling out of use, the 1SG and 3PL are likely to have influenced one another and to have become just sou in later Britainese (for 1SG and 3PL being identical, cf. Romanian sunt, Italian sono, Corsican sò).

The Romance 2SG and 3SG are derived from Latin es [ęss] and est [ęst]. In some areas there was a remodelling to have all forms beginning with [s] so we find, e.g. Italian sei, Corsican sì; and we find 3SG as è in Italian and Romansh, and e (besides este) in Romanian which is possibly from a remodelling of Vulgar Latin es, *e by analogy with *as, *a, see 9.3.4.2 below (the reduction of French est to [ę] is not related and developed in French as final cosonants fell silent). It clear from the table above these remodellings did not affect the langues d'oïl and there is no reason to suppose they would have affected early Britainese where es(s) and est would have persisted, giving was in later Britainese to ess for both persons.

With the IPL and 2PL we see a great variation in the Romance languages, a variety that dates back at least to the time of the Emperor Augustus who, were are told by Suetonius, pronounced the 1st plural as simus ['simʊs], a form that endured in the Vulgar Latin ['sẹmọs] of southern Gaul and Italy. There was also remodelling as one form influenced the other. Thus we find Vulgar Latin:

• sumus ['sọmọs], estis [ęstẹs]
• sumus ['sọmọs], *sutis ['sọtẹs]
• simus ['sẹmọs], *sitis ['sẹtẹs]
• *esmus '[ęzmọs], estis [ęstẹs]

The table above might suggest that the langues d'oïl all knew only (i) above; but that is not so. Certainly (ii) and (iii) are not found in north France; but (iv) was found as the Old French 1PL was esmes or somes; the latter sommes is a conflation of the two (somes would have become *sons.

Clearly the Norman form is of the same origin. The Vulgar Latin [ęstẹs] would have become ests in Britainese which has an awkward combination of consonants and in my opinion would not have been stable. It is likely early Britainese also knew a 1PL esms. But of interest from the table above is that Gallo has alternastve forms derived from stare and in Walloon they are the only forms. Britainese would have known these also and those stable forms will have ousted the unstable forms derived from (i) and (iv) above.

BRITAINESE
Infinitive	star "to be"
eu sou   tu ess   il/el ess	nos stams vos stadz ils/els sou

9.3.4.2 TO HAVE

While the verb "to be" tends to be very anomalous, e.g. English am, is, are with archaic and poetic art with what are synchronically three different stems, "to have" usually shows just a single basic stem with more minor irregularities e.g. English which has have but but not *haves, that being replace by has. Cf. the table below:

FRENCH	GALLO	NORMAN	PICARD	WALLOON	OCCITAN
avoir	aveir	aveir	avoèr	avu	aver
j'ai tu as il/elle a nous avons vous avez ils/elles ont	j'ae t'as il/al at j'avom voz avaez il/al ont	j'ai t'as il/ol a nos avoms vos avez il/ol ont	j'ai t'os il/al o (n)os avons* os avez* is ont	dj'a t'as il/ele a nos avans vos avoz i/ele ont	ai as a avèm avètz an

* Some Picard dialects have contracted forms of 1PL and 2PL: ons, ez.
The 1PL and 2PL are occasionally found contracted in other Romance languages or dialects; but the,vast majority, as in the table above, form these from Latin habēmus and habētis; Britanese does likewise.

The 1SG, 2SG, 3SG & 3PL developed contracted forms in Vulgar Latin thus:

• 1SG *ajo ← *aβjo ← hábeō or *ajo ← *aβo ← hábeō
• 2SG *as ← *aβs ← hábēs
• 3SG *a(t) ← *aβt ← hábet
• 3PL*an(t) or *áun(t) ← *aβnt ← hábent

Occitan an "they are" is from •an(t); ont in French and other langues d'oïl may be from *aunt or a remodelling of *ant influenced by sont "(they) are."

Early Britainese knew ai [ai̯] or ae ['a.ə], as, ad and ant or aunt. The singular would all eventually have become á for reasons explained in 9.3.3.1 above. When 3PL final -nt was lost this would also become á (The acute accent on á is to distinguish it from the unstressed preposition a, adh "at, to.").

BRITAINESE
Infinitive	avair "to have"
eu á   tu á   il/el á	nos avams vos avadz ils/els á

9.3.4.3 TO DO

In English this verb, like "to have" is only slightly irregular; we have do [duː], but not *dos [duːz]. Instead the 3SG is does [dʌz]. Brittainese has a similar pattern to thast of "to have" above, but the route is more complicated because of the variety of developments in Proto-Romance.

The languages shown are different from those in the two sections above as I do not know the forms for the other langues d'oïl. I give Occitan, the Romance language of southern France, and two neighbouring Romance languages - Catalan to the south and Romansh to the east.

FRENCH	WALLOON	OCCITAN	CATALAN	ROMANSH
faire	fé	faire	fer	fare
je fais * tu fais il/elle fait nous faisons † vous faites ils/elles font	dji fwai ti fwais i/ele fwait nos fijhans vos fijhoz i/ele foaiynut	fau fas fa fasèm fasètz fan	faig [fat͡ʃ] fas fa fem feu fan	jau fatsch ti fas el/ella fa nus faschain vus faschais els/ellas fan

* In Old French faz [fat͡s] (← facio); fac [fak] and fai [faj] (← *faco)
† In Old French faimes

Besides the Classical imperative fac, the form fa is attested (Zeitschrift für romanische Philologie, Halle, XXV, 735), which is likely to have led to plural *fáte besides the Classical facite. From this and under the influence of stāre and dare, *fāre as a doublet of facere existed, with present tense:
*fo, *fas, *fat, *fámus, *fátis, *fant.
From these forms are clearly derived 2SG, 3SG and 3PL of Occitan, Catalan and Romansh. Although it is tempting to have Britainese derive its verb "to do" from this or, partly from this, it does seem to have been a later developed beginning in Italy and spreading to adjacent areas (see the Wiktionary entry); although it reached southern Gaul it does not seem to reached northern Gsul and, therefore, unlikely to have reached Britain.

Latin facio, facere developed variously in Vulgar Latin:

1. As one would expect (see section 9.1 above): facio ['fakʲọ], facis ['fakʲẹs], facit ['fakʲẹt], *facémus [fa'kʲẹmọs], *facétis [fa'kʲẹtẹs], faciunt ['fakʲọnt]
2. As we saw in section 9.3.1 above, 1SG ending -io and 3PL -iunt generally simply become [ọ] and [ọnt] in Vulgar Latin, i.e *faco ['fakọ] and *facunt ['fakọnt]
3. In many areas the unaccented /i/ before consonant endings was dropped and we find contracted forms for 2SG, 3SG, 1PL, 2PL, thus: *facs [faks], *fact [fakt], *facmus ['fakmọs], *factis ['faktẹs]

We see (i) in the 1SG of Catalan and Romansh and Old French faz; and in the 1PL and 2PL of Walloon, Occitan and Romansh. The other forms do not seem to have been productive (Iberian Romance went its own way and assimilated the verb to the Latin 2nd conjugation, i.e. their 3PL is from *facent [['fakʲęnt]).

That (ii) occurred is shown clearly in Romanian 1SG and 3PL fac [fak]. As for languages in table above, Old French fac and, probably, fai and Walloon fwai are from *faco; the French 3PL may be from */'faɣọnt/ ← */'fagọnt/ ← */'fakʊnt/ or simply a remodelling of *fant under the influence of sont "[they] are." Walloon foaiynut is a later internal development which does not concern us here.

In the contracted forms of (iii) the pre-consonantal [k] will have become [j] or [i̯] in the Romance languages. It is from the contracted forms that the 2SG, 3SG, 1PL and 2PL of Old French and of Catalan are derived (the modern French faisons is a later remodelling in French), and the 2SG and 3SG of Walloon.

From the above, it seems most likely that in early Britainese:

• 1SG and 3PL had forms derived from (ii) with 1SG having an alternative derived from (i), thus fag [fag] or fage [fad͡ʒ], and fagnt [fagn̩t];
• all the others derived from (iii) with alternatives for 1PL and 2PL from (i), thus fais [fai̯s], fait [fai̯t], faims [fai̯ms] or fageams [fa'd͡ʒamz], faits [fai̯ts] or fageadz [fa'd͡ʒadz].

In middle Britainese there would be moves towards standardization. Almost certainly the anomalous faims and faits would have given way to fageams, fageadz; fais and fait would have become just fai, putting pressure on the 1PS to fall in line and also become fai; but the 3PL fagnt will surely have given support to fag, especially when final -nt fell silent. Thus we would have a paradigm reminiscent of "to be" above where 1SG and 3PL are the same, 2SG and 3SG are the same, and 1PL and 2PL have related forms with regular endings. But the verb "to be" will be more common than "to do" and while sou and ess are quite distinctive, fag and fai are not. The pressure from 2SG and 3SG will be too great for 1SG to resist and fag will give way to fai both there and eventually in the 3PL.

BRITAINESE
Infinitive	fair "to do"
eu fai   tu fai   il/el fai	nos fageams vos fageadz ils/els fai

9.3.5 Some other anomalous present tenses

This is not intended to be a complete list of irregular present tenses, but indicative of the sort of irregularities, other than those with change of stem vowel in 1st & 2nd persons plural (see 9.3.3 above) that may occur. Indeed. there will be fewer irregular present tenses than there are in French as the latter language:

1. had a secondary lenition whereby, e.g. both intervocalic /b/ and /p/ become /v/, whereas Britainese had only primary intervocalic lenition, e.g. /b/ → /v/, and /p/ → /b/.
2. both [ð] and [ɣ] fall silent, whereas only [ɣ] fell silent in Britainese which retained [ð].

This has meant that Vulgar Latin intervocalic /d/, /g/, /t/, /k/ have all fell silent in Old French, leading to awkward hiatus which was resolves in various ways. In Britainese only Vulgar Latin /g/ became silent in Old Britainese.

This means, e.g. that the verb "to be able" has a regular present tense in Britainese. The highly irregular Classical Latin possum, posse, potuī was replaced in Vulgar Latin by poteo, potére, potui; notice that the Vulgar Latin forms are not preceded by asterisks as they are actually attested in written Latin during the 1st millennium AD. The Old 1st person singular did survive for a time as posso (and Old French puis suggests a form *possio also), but the 1st person singular pot(e)o seems to have prevailed in most places. The present tense of Britainese podair is formed regularly with stem pod.

I give two verbs below as indicative of the sort of irregularities that might occur.

9.3.5.1 ALAR "to go"

The Classical Latin for "to go" is īre which my headmaster of 70+ years ago used to call a "Cheshire Cat verb" because "the stem had disappeared, leaving only the endings." This is not strictly true, but it certainly looks like it. It seems that of the present tense only 1PL īmus and 2PL ītis survived in Vulgar Latin with suppletion from contracted forms derived from Latin vādere "to go, to walk.", namely: 1SG *vo ← vádo, 2SG vas ← *vads ← vadis, 3SG *vat ← *vadt ← vadit, 3PL •vant ← *vadnt ← vádunt. The 1PL also had *vajo formed by analogy with *ajo "I have". This remained in the Iberian peninsula but modern Portuguese has replaced the 1PL with vamos, still retaining 2PL ides, while modern Spanish now has 1PL and 2PL vamos, vais.

In Italy and Gaul īmus and ītis were replaced by verb stems whose origins are still disputed:

• anda- in Italy;
• an(n)a- in southern Gaul and Catalonia;
• al(l)- in northern Gaul.

Thus:

FRENCH	WALLOON	OCCITAN	CATALAN	ITALIAN
aller	aler	anar	anar	andare
je vais tu vas el/ille va nous allons vous allez ils/elles vont	dji va ti vas i/ele vat nos alans vos aloz i/ele vont	vau vas va anam anatz van	vaig vas va anem aneu van	vado, vo* vai va andiamo andate vanno

* vado is standard; vo is archaic and poetic.

At one time the verbs andare, an(n)ar and al(l)er were all held to be derived in some peculiar way (distortion of military commands was often suggested) from Latin ambulāre "to walk." But I have never been convinced by this; it is more likely that:

• andáre, attested in Medieval Latin (in a text from southern Italy around 800 AD) is from *ambitāre "to go around", being formed from ambitus "circuit, orbit." From it also are derived the Spanish and Portuguese andar "to walk" (quite distinct from their verbs ir "to go");
• Old Occitan annar and modern Occitan and Catalan anar are possibly derived from *amnáre ← *amláre ← *ambláre (Classical ambulāre);
• The origin of French all- (old French al-) is obscure. A Medieval Latin aláre is attested in the Reichenau Glossary, a collection of Latin glosses compiled around the eighth century at the Abbey of Corbie in Picardy. It has been suggested that it derives it from Gaulish *aliu, from Proto-Celtic *ɸal- or *ɸel- "to approach, drive" ← Proto-Indo-European *pelh₂- (compare Welsh elwyf "I may go", Cornish ellev "I may go").

It will be seen from the above that andare, with the meaning "to go", is confined to Italy; an(n)ar is found in southern Gaul and Catalonia; alar is attested in Medieval Latin in a glossary written in Picardy and was clearly known in northern Gaul. If the Celtic origin al- is correct then this verb must surely have been known in Britain also. So in Britainese we find:

BRITAINESE
Infinitive	alar "to go"
eu va   tu va   il/el va	nos alams vos aladz ils/els va

9.3.5.2 DIR "to say, to tell"

All the Romance languages derive their verb "to say" from Latin dīcō, dīcere, with forms derived from [dik] before central and back vowels and from [dikʲ] before front vowels. In some areas, we find forms for all except 1PSG and 3PL derived from contractions: *dics ← dīcis, *dict ← dīcit, *dícmus ← dīcimus, *díctis ← dīcitis

FRENCH	WALLOON	OCCITAN	CATALAN	ITALIAN
dire	dîre	dire	dir	dir
je dis tu dis el/ille dit nous disons * vous dites ils/elles disent	dji di ti dis i/ele dit nos dijhans vos dijhoz i/ele dijhèt	disi dises ditz disèm disètz dison	dic dius diu diem dieu diuen	dico dici dice diciamo dite dicono

* In Old French dimes.

In early Britainese this would have given:
dig, did, dit, dims or digeams, dits or digeadz, dignt. Which, for similar reasons to those discussed under "to do" above would have given:

BRITAINESE
Infinitive	dir "to say"
eu di   tu di   il/el di	nos digeams vos digeadz ils/els di

Note:
Latin verb dūcere "to lead, guide" gave Old Britainese eu dug, tu du(i)s, el/il du(i)t, nos dugeims/dugeams, vos dugeids/dugeadz, els/ils dugnt which was all very well when u was pronounced [y]; but when the vowel became unrounded the verb became homophonous with dir "to say"! The verb dropped out of use, just as it did in French and, indeed, most western Romance languages, surviving only as the second element of compound verbs, cf. French conduire, produire, induire, Spanish conducir, producir, inducir (but no verb *ducir), Portuguese conduzir, produzir induzir (but no verb *duzir). In Catalan there is a simple verb dur "to carry, bring", but when it occurs as the second part of conduir, produir, induir it has not only a different infinitive, but also becomes an inchoative -ir verb. Similarly in Britainese the verb was remodelled as the second part of condugir, produgir, indugir, being reformed on the 1PL ans 2PL -dug(e)- [did͡ʒ] and the infinitive ending -ir.

9.4 The Imperfect Indicative Tense

The Imperfect is the 'past of the present', i.e. just as the simple present tense in Britainese denotes an habitual action or state in present time, so the Imperfect Indicative denotes habitual action or state in past time, i.e. el cantai = " she sang [habitially], she used to sing.". The meaning "she was singing" is expressed by a periphrastic construction similar to English or Welsh in our timeline. In this respect, cf. the table below:

Language	Past Continuous	Past Habitual	Preterite
Welsh	roedd hi'n canu	fe ganai hi	fe ganodd hi
English	she was singing	she sang	she sang
French	elle chantai	elle chantai	elle chanta

For more on the Preterite (i.e.Past Perfective), see section 9.5 below and, indeed, compare the table above with a similar one of the Preterite and the Present Perfect given in that section.

From the table above it will be seen that while in Welsh we have three separate tenses, in English the Simple Past may be used both for habitual past actions and for the past perfective; but past continuous is shown by by was/were and the imperfective participle. Whereas in French the Imperfect does duty both for past continuous and the past habitual, the Passé Simple denotes the past perfective (though much restricted in modern French, where the Passé Composé normally does duty as both a present perfect and a past perfective). Britainese has a Preterite, as we shall see below, and the past continuous is shown by "to be" and the gerund. The imperfect, therefore, with dynamic (or fientive) verbs has a past habitual meaning (we shall discuss stative verbs when we look more fully a verbal syntax).

9.4.1 Personal endings

Latin had the same personal endings for all verbs; these were, except for the 1st person singular, identical with the personal endings of the present 1st conjugation. The Britainese forms are, therefore, as the present 1st conjugation forms above, namely:

	1SG	2SG	3SG	1PL	2PL	3PL
Latin	-a(m)	-ās	-at	-āmus	-ātis	-ant
Early Britainese	-e	-es	-et	-ams	-adz	-ent
Late Britainese	-	-	-	-ams	-adz	-

9.4.2 Temporal suffix

Between the verb stem and the personal endings, Classical Latin had the following temporal suffixes:

• 1st conjugation -āb-
• 2nd and 3rd conjugations (except "faciō" verbs) -ēb-
• 3rd conjugation "faciō" verbs and 4th conjugation -iēb-

In the 4th conjugation in early Latin the suffix was -īb-, and this reappears in Late Latin. It seems to have been the norm in spoken Latin. In Vulgar Latin also, the suffix -iēb- became just -ēb-. Therefore early Vulgar Latin had the temporal suffixes:

• -aβ- verbs with infinitives in -áre
• -ẹβ- verbs with infinitives in -ére or -ere
• -iβ- verbs with infinitives in -íre

In western Romance, while -aβ- remained, -ẹβ- became -ẹ- and -iβ- became -i-. It is not clear why [β] disappeared after [ẹ] and [i]. Grandgent suggests that it began with habēbam [a'βẹ.βa] (imperfect of "to have") becoming *[a'βẹ.a] through dissimilation and that affected all except -aβ-. Furthermore, both -ẹ- and -i- fell together, with -i- being normal in southern Gaul and the Iberian peninsula, and -ẹ- prevailing in northern Gaul.

In Old French we find that:

• the first conjugation -aβ-, when stressed became -év- in some eastern regions, in others became -ou- or -o- (← -aw- ← -aβ-); and that when unstressed before 1PL and 2PL endings it became -i- everywhere.
• the -ẹ- of other verbs became -ei- → -oi- when stressed, and -i- when unstressed.

In French -i- before the 1PL and 2PL endings has remained till the present day; at an early date -oi- replaced the 1st conjugation -ou-, -o- and developed [oi̯] → [wę] → [ę], the spelling oi starting to change to ai in the 17th century but not recognized by the French Academy till 1835.

There can be little doubt that Britainese would also have had two temporal suffxes:

• Verbs with infinitives in -ar would have had -av- (or -au-);
• all other verbs (except "to be", see below) would have had -ei- → -ai-;
• the verb "to be" in early Britainese would have er- besides the regularly formed imperfect of star.

Would both the -av-/-au- and the -ei- → -ai- imperfects fallen together as they did in France? According to Meyer-Lübke it was in southern Wallonia, Picardy, the Ile-de-France, Franche-Compté, Lorraine and Burgundy where the -oi- imperfects early ousted the -ev-, -ou- imperfects. Part of this area is too far to the south-east for us worry about, but southern Wallonia and Picardy are close to the Britainese area. Also if we look at our 'control languages' we find that both Welsh and French in our timeline have only one set of endings for the imperfect (English does not have a synthetic imperfect and its Simple Past tense is more comparable to Britainese Preterite which we shall consider later).

Therefore, if we follow the principles on which we have evolved Britainese, we must accept that in Britainese the -ai- imperfect endings will have ousted the -av-, -au- endings - probably at first in the south east and then gradually spreading to all areas during the Old and Middle Britainese periods. I do not, however, think the temporal suffix would have weakened to -i before the 1st and second person plural endings; that would, among other things, have involved palatilization of dental/alveolar and velar consonants. I am sure it would just have been a weakened, unstressed -ai- [əj], e.g. nos cantaiams [nus'kən'tə'jamz] "we used to sing."

As regards the '-ir inchoative verbs', it will be found that those Romance languages in which the inchoative suffix is kept between the verb stem and personal endings throughout the present tense also retain the inchoative suffix in the imperfect tense. It would be odd if Britainese were to be an exception. Therefore the imperfect indicative of, e.g. flourir is eu flourissai.

9.4.3 Verb "to be"

The Vulgar Latin verb essere (like Clasical Latin esse) "to be"; used the same imperfect indicative personal endings as all other verbs, but it had no temporal suffix; in both Vulgar Latin and the Classical language, these endings were suffixed directly to the stem er- thus: era(m), eras, erat, erámus, erátis, erant. Forms derived from these were found in all Romance languages, including Old French and related langues d'oïl; but in French and related languages they have yielded ground to forms which, which, according to Meyer-Lübke, were derived by abstracting est- from the infinitive estre (← essere) and adding the imperfect endings -oie, -oies, -oi(t) &c. imperfect endings. Others take it as the imperfect of ester( ← stáre), but the expected imperfects *estoe, *estoie, estou(t) &c. are not found; maybe it was a confusion of the two verbs that gave rise to estei-, estoi- which were actually found, from which are derived the the modern étai-.

We find similar developments in Bourguignon (étoos, étoos, étoot ..., Franc-comtois (étavo, étaves, étave ...), Gallo (estei, esteis, esteit ...), Lorraine (estoi, estois, estoit ... and other variants), Norman (éteis, éteis, éteit ...), Picard (étoé, étoés, étoét ...) and Walloon (esteu, esteus, esteut ... ); though Franc-Comptois and Lorrain also have alternative imperfects derived from Latin era(m), eras, erat ... and Walloon also has alternative forms derived from Latin era(m), eras, erat ... in the singular only.

Clearly this was a development going on the the northern part of the Romance speaking area. Although forms from the Latin survived in as alternatives in Franc-Comptois and Lorrain and, to a more limited extent, in Walloon, it is noteworthy that forms begining e(s)t- are found in all these languages and that in Norman and Picard they are the only forms found. This would surely suggest that although forms such as ere ['erə], eres ['erəz], ered ['erəd] &c. survived in Old Britainese, they gave way to stai, stai(s), stai(d) &c., i.e. st- abstracted from the infinitive with the imperfect endings of all other verbs.

9.4.4 Britainese examples

star "to be"	cantar "to sing"	vendr "to sell"	dormir "to sleep"	finir "to finish"
eu stai tu stais el/il stai nos staiams vos staiadz ils/els stai	eu cantai tu cantai el/il cantai nos cantaiams vos cantaiadz ils/els cantai	eu vendai tu vendai el/il vendai nos vendaiams vos vendaiadz els/ils vendai	eu dormai tu dormai el/il dormai nos dormaiams vos dormaiadz els/ils dormai	eu finissai tu finissai el/il finissai nos finissaiams vos finissaiadz els/ils finissai

vedhair "to see"	baivr "to drink"	alar "to go"	avair "to have"	fair "to do"
eu vedhai tu vedhai el/il vedhai nos vedhaiams vos vedhaiadz ils/els vedhai	eu bevai tu bevai el/il bevai nos bevaiams vos bevaiadz ils/els bevai	eu alai tu alai el/il alai nos alaiams vos alaiadz els/ils alai	eu avai tu avai el/il avai nos avaiams vos avaiadz els/ils avai	eu fageai tu fageai el/il fageai nos fageaiams vos fageaiadz els/ils fageai

9.5 The Preterite Indicative Tense

The Preterite Tense of the Romance languages is derived from the Latin Perfect Tense; that tense in Latin functioned both as a present perfect, i.e. to express a past event that has present consequences as in "I have arrived", and as a past perfective tense, i.e. a past event viewed as a single whole* as "I went there yesterday."

*It is important not to confuse perfective aspect, event/action viewed as a single whole, with the similarly named perfect, a state resulting from a previous event/action.

We saw above from subsections 9.2.1.1, 9.2.2.5(iii) and 9.3.4 that the Latin present perfect was expressed in Vulgar Latin by

• "to be" with perfect participle agreeing with the subject if the verb was intransitive, or
• "to have" with perfect participle agreeing with the direct object if the verb was transitive;

and that the Classical Latin Perfect had become the past perfective, or Preterite Tense.

At least that is so in theory, just as it is with the English 'Simple Past' (in theory, past perfective) "I went" and 'Present Perfect' "I have gone." But in practice there is overlap between the uses of these two tenses; and usage differs within the English speaking word. I recall several years ago my daughter, who has lived in the USA for several decades, married an American and has US citizen, asked me: "Did you go to the shops yet?" which took me surprised because I would have said "Have you been to the shops yet?". We find quite an overlap between the two tenses in the Romance languages as the names given to the tenses in all, except Spanish, clearly suggest:

Language	Simple Past (Preterite)	(Present) Perfect
French	passé simple	passé composé
Portuguese	pretérito perfeito	pretérito perfeito composto
Spanish	pretérito indefinido	pretérito perfecto
Italian	passato remoto	passato prossimo
Romanian	perfectul simplu	perfectul compus

It is worth reading the Wikipedia article "Preterite", especially the section on Romance languages. It will be seen that the preterite (passé simple) has disappearead from spoken French and in the written language is retain retained only in narrative. In Romanian and Italian use varies from region to region. It is tempting to say the tense died out in Britainese but Britainese would have had this tense in its earlier history and the evidence of our 'controlling languages' in our timeline suggests that it is likely to have survived in Britainese, cf.

Language	Preterite	(Present) Perfect
Welsh	fe fytodd e	mae e wedi bwyta
English	he ate	he has eaten
French	il mangea	il a mangé

Of those three languages, only French will exist in BART where Britainese will have il mangá and il á mangad respectively.

9.5.1 Preterites in Vulgar Latin

Those who have learnt Classical Latin will know that the Perfect tense had the personal endings: -ī, -istī, -it, -imus, -istis, -ērunt and that these were added to the perfect stem of all verbs without exception. They may also remember that in the 2SG, 2PL and 3PL contractions occurred, particularly in verse, e.g. amāstī ← amāvistī, amāstis ← amāvistis, amārunt ← amāvērunt; fīnīstī ← fīnī(v)istī, fīnīstis ← fīnī(v)istis, fīnīrunt ← fīnī(v)ērunt.. These contractions reflect what was happening in the spoken language. The Vulgar Latin Preterites are usually divided into Weak, of which there were three variants, and Strong, of which there were also three variants.

9.5.1.1 Vulgar Latin Weak Preterites

The common feature of weak preterites is that they keep the stress uniformly on the first vowel of the endings. They are the 1st conjugation amái (Classical amāvī), 4th conjugation finíi (Classical fīnīvī) and a formation peculiar to Vulgar Latin, e.g. *vendę́di ← véndere "to sell" (Classical Latin vēndidī /ˈu̯eːn.di.diː/ ). Note that the Vulgar Latin form has the stress shifted to the ending and has a low-mid [ę] and not the expected high-mid [ẹ] from Latin [ɪ]; these arose from a transformation of -didī under the influence of dedī, the reduplicated preterite of dare "to give." Such forms as credę́di, perdę́di are attested in Late Latin texts (without, of course, the diacritics) and, indeed, extended to other verbs whose preterites did not end in -didī in Classical Latin, e.g. respondę́di ← respondére "to reply", defendę́di ← deféndere "to defend", cadę́di ← cádere "to fall".

Also we should add that the very few Classical preterites of the form dēlēvī ← dēlēre "to destroy" did not survive in Vulgar Latin, giving way to preterites ending in -íi.

The weak preterites of Vulgar Latin were:

1SG	cantái	finíi	vendę́di
2SG	cantásti	finísti	vendísti
3SG	cantát, cantáit, cantáut (← amávit)	finít, finíit finíut	vendę́dit
1PL	cantámus, cantáimus	finímus, finíimus	vendę́dimus, vendímus
2PL	cantástis	finístis	vendístis
3PL	cantárunt	finírunt	vendę́derunt, vendę́runt

9.5.1.2 Vulgar Latin Strong Preterites

These retained the old Latin 3PL ending -erunt (not the Classical -ḗrunt); this form was still encountered in verse in the Classical period and clearly remained the spoken form. The Strong Preterites differ from the Weak in that the stress is on the verb stem in the 1SG, 3SG. 1PL and 3PL but on the suffix in 2SG and 2PL; but the stress shifted to the suffix in the 1PL in some areas, under the influence of finímus &c.

Classical Latin had five Strong variants:

1. reduplicated, e.g. momórdī ← mordḗre "to bite";
2. those that add personal endings directly to stem (sometimes with modification of stem vowel), e.g. vḗnī ← venī́re "to come";
3. those adding -s- to the stem, e.g. scrī́psī ← scrī́bere, "to write";
4. those adding -u- to the stem, e.g. sápuī ← sápere/sapḗre "to know [a fact], know [how to]";
5. the verb "to be" which had fúī [ˈfʊ.iː] or fū́ī [ˈfuːiː].

Of these, (i) did not survive in Vulgar Latin. Many originally in group (ii) and some in group (i) gave way to (iv), e.g. bíbi "I drank", crédidi "I believed", cécidi "I fell", cúcurri "I ran",féfelli "I deceived", recépi "I received", légi "I read" gave way to *bíbui. *crédui, *cádui, *cúrrui, *fállui, *recépui, *légui respectively. According to Meyer-Lübke, Grammatik der romanischen Sprachen II p.357, preterites in -v- not made from the present stem, e.g. cognōvī ← cognōscere, crēvī ← crēscere, mōvī ← moveō, pāvī ← pāscere were also part of group (iv), the written v being [wu], i.e. mōvī ['mo:wu.i:].

Some other preterites of groups (i) and (ii) joined group (iii), either under the influence of the present indicative, e.g. áttigi "I touched", frégi "I broke", púpugi "I pricked" gave way to *attánxi, *franxi, *púnxi, or under the influence of their perfect participle, e.g. elégi "I chose", momórdi "I bit", occídi "I slew", prehéndi "I seized", redémi "I redeemed", sédi "I sat" gave way to *eléxi, *morsi, *occísi. pré(n)si, *redémpsi, *séssi.

Of group (iv), when the ending is stressed, the u fuses with the ending as in the Weak forms above, e.g. sapuisti → *sapústi; but when the stress falls on the verb stem, then:

• after a group of consonants or a geminated consonant, or unstressed u became [w] and fell silent or was lost by dissimilation after initial v [w], e.g. válui → váli, and joined (ii);
• after a single consonant other than a liquid or nasal, it was attracted to the preceding stressed vowel, e.g. sapui → *saupi;
• after a liquid or nasal it remained stressed and thus the verb behaved like a Weak Preterite, e.g. parúi (← parére "to appear, be apparent"), see below.

The preterite of "to be" had stress on the first syllable in Classical and Vulgar Latin. In Classical Latin the stem was fŭ- with a short vowel; but in early Latin the stem was fū- with a long vowel; such forms are found in Ennius and Plautus. Possibly the practice in Vulgar Latin was regional (or even individual); the forms given in the table below are based on those given by Grandgent (1907).

Examples of Vulgar Latin Strong Preterites (and one more Weak Preterite) and the verb "to be":

	Group (ii) (Strong)	Group (iii) (Strong)	Group (iv) (Strong)	Group (iv) (Weak)	"to be"
1SG	véni	scrípsi	sáupi	parúi	fúi, fó̩i
2SG	venísti	scripsísti	sapústi	parústi	fó̩sti
3SG	vénit	scrípsit	sáupit	parú(i)t	fó̩(i)t, fú(i)t
1PL	vénimus, venímus	scrípsimus, scripsímus	sáupimus, sapú(i)mus	parú(i)mus	fó̩m(m)us
2PL	venístis	scripsístis	sapústis	parústis	fó̩stis
3PL	vénerunt	scrípserunt	sáuperunt	parúrunt	fó̩runt, fúrunt

9.5.2 Preterites in Britainese

Although vendedd was found in some early Britainese texts, because of Vulgar Latin vendísti, vendímus, vendístis the verb and other similar verbs were attracted into the finíi group. It meant that in early Britainese there were three weak preterites (stressed throughout on the the suffix). Indeed, it will be found that there has been considerable configuration of this tense in the various Romance languages and dialects. The examples of Romance Preterites in the Wikipedia article on Preterites give some small indication of this. I will not go into all the many other variations here, but consider mainly at what went went on in Gaul and what might have happened in Britain.

Of the langues d'oïl I have considered in sections above, Picard has dropped this tense entirely, and Walloon uses a single set of endings for all verbs: -a, -as, -a, -îs, -îz, -ît. Although, as I have observed above, it would be tempting to follow Picard, I think the tense would have survived in Britainese. We have seen, however, that there was a strong tendency in Britainese towards having a single set of endings for all verbs in finite tenses so it is likely that Britainese will follow Walloon in that respect.

Walloon's three plural endings are odd; but it will be noticed that Latin -astī is just -as. Indeed, only Italian preserves the Latin ending, while -stī survives as -ste in Iberian Romance. Throughout Gaul in both the langues d'oïl, Franco-Rovençal and the Occitan dialects the final -tī has vanished without trace and this was likely so in Britain also. It would seem that in southern Lorraine, Wallonia and Picardy and probably other areas, the endings -ai, -as, -a(t) of verbs like c(h)antar came to be influenced by the Present of the verb "to have", leading in some cases to the 3PL of the Preterite being giving the ending -ont (in Britainese that would be -ant). I find no reason to suppose that something similar went on in early Britainese. As the 1Sg, 2Sg, 3SG and 3PL of "to have" came to have a single form á so these persons of the Preterite of cantar came to be cantá and also that this ending gradually came to spread to other week Preterites as the 'normal' ending.

The singular of strong preterites would have had no ending for 1SG, -s for 2SG and an unstable -d or -t in the 3SG. Eventually we can expect all singular persons to have no endings which would have put pressure on the 3Pl -rnt; as the 3PL -nt was falling silent generally the 3PL would be left with [r̩] which would have eventually dropped under the influence of week Preterite 3PL having the same form as the three singular persons, e.g. vain(t)r → vain like 1SG, 2SG & 3SG vain.

The strong preterites would have had 1PL in -ims and 2PL in -ists; but although in later Britainese the combination [sts] would have been encountered in plurals such as tests "heads" for earlier testes ['testəz], the singular would support the retention of [st] and the final [s] was the plural morpheme, i.e. final [sts] occurred only where final [s] was a separate morphemes; that would not have the case with Preterite 2PL -sts which would have been unstable. We find this as -stt in Gallo and it similarly simplified to -ist in the early Britainese period.

The occurrence of these endings with strong preterites and their occurrence with most verbs other than the cantar led these endings spreading to the latter. Indeed, cantams which was identical with the present tense 2PL readily gave way to the more distinctive cantims and eventually cantast(s) gave way to cantist. Thus by the middle Britainese period we have a single set of endings for week preterites and another for strong preterites (which included the preterite of the verb "to be" and other auxiliaries). Thus we find by late Britainese:

• Week Preterites:

  cantar	finir	vendr	parair
  eu cantá tu cantá el/il cantá nos cantims vos cantist els/ils cantá	eu finá tu finá el/il finá nos finims vos finist els/ils finá	eu vendá tu vendá el/il vendá nos vendims vos vendist els/ils vendá	eu pará tu pará el/il pará nos parims vos parist els/ils pará

• Strong preterites:

  venir	scrivr	avair	fair	star
  eu vain tu vain el/il vain nos vainims vos vainist els/ils vain	eu scriss tu scriss el/il scriss nos scrissims vos scrissist els/ils scriss	eu auv tu auv el/il auv nos auvims vos auvist els/ils auv	eu faige tu faige el/il faige nos faigims vos faigist els/ils faige	eu fou tu fou el/il fou nos föims vos föist els/ils fou

Note: föims = [fu'imz] and föist = [fu'ist]

9.6 The Future and Conditional Tenses

The Classical Latin futures did not survive in Vulgar Latin except in northern Gaul for the verb "to be." Instead it gave way to a variety of periphrases; but the only one that concerns us here is the infinitive with the verb "to have" which became the norm in western and Italian Romance, e.g. cantāre habeō "I have to sing" → "I shall/will sing." In all the western Romance languages the two parts fused, e.g.

• French: je chanterai (je + chanter + ai) ← chanter "to sing" + ai "I have"
• Occitan: cantarai (cantar + ai) ← cantar "to sing" + ai "I have"
• Portuguese and Galician: cantarei (cantar + [h]ei) ← cantar "to sing" + hei "I have"
• Spanish and Catalan: cantaré (cantar + [h]e) ← cantar "to sing" + he "I have"
• Italian: canterò (cantar + [h]o) ← cantare "to sing" + ho "I have"

and thus:

• Britainese: eu cantará (eu + cantar + á) [əu̯.kən.tə'ra] ← cantar "to sing" + ai "I have"

This future form also has a past tense equivalent formed from infinitive and the imperfect indicative of habeō, i.e. cantāre habēbam "I had to sing" → "I would sing." It was used like the 'Future in the Past' in English to show a shifted future in reported speech after a past main verb, cf:

• "I will go" (Direct Speech)
• She said she would go. (Reported Speech)

As with the English 'Future in the Past', this is also used in some types of conditional sentences (though not in every sort of conditional sentence), e.g. "If he went there, he would see it for himself." Because of of this, it has traditionally become known as the Conditional Tense in Romance languages.

You may notice that I did not write "The Future and Conditional Indicative Tenses" above. Some, with good reason in my opinion, consider the Conditional to be a different Mood, i.e. neither Indicative nor Subjunctive. In some analyses of Romance verb, the Future and Conditional are listed as Tenses of the Potential Mood.

9.6.1 Future Tense

As we saw above, the future indicative is formed from a fusion of "to have" and the infinitive; however, in common with other related Romance languages, the 1st and 2nd persons plural will have shortened forms, namely just -ams and -adz in Britainese, thus:

cantar "to sing"	alar "to go"	vendr "to sell"	dormir "to sleep"	finir "to finish"
eu cantará tu cantará el/il cantará nos cantarams vos cantaradz ils/els cantará	eu alará tu alará el/il alará nos alarams vos alaradz ils/els alará	eu vendrá tu vendrá el/il vendrá nos vendrams vos vendradz ils/els vendrá	eu dormirá tu dormirá el/il dormirá nos dormirams vos dormiradz els/ils dormirá	eu finirá tu finirá el/il finirá nos finirams vos finiradz els/ils finirá

In verbs whose present tense shows apophany, the unstressed stem of the 1st and 2nd persons plural is used; thus the future of boivr "to drink" has the stem bevr-. Also verbs whose infinitives end in -air form futures as though the infinitive ended in -r; this will sometimes mean that an epenthetic consonant appears between the verb stem and -r-, e.g. the future of volair "to wish, want" has the stem voldr-; fair [fəi̯r] was weakened to [fər] rather than [fr̩]; this was spelled as fer- or far-, but far- became standard in later Britainese.

baivr "to drink"	vedhair "to see"	avair "to have"	volair "to wish, want"	fair "to do"
eu bevrá tu bevrá el/il bevrá nos bevrams vos bevradz ils/els bevrá	eu vedhrá tu vedhrá el/il vedhrá nos vedhrams vos vedhradz ils/els vedhrá	eu avrá tu avrá el/il avrá nos avrams vos avradz ils/els avrá	eu voldrá tu voldrá el/il voldrá nos voldrams vos voldradz els/ils voldrá	eu fará tu fará el/il fará nos farams vos faradz els/ils fará

There remains to consider the verb "to be." In Old French we find three different formations for the future:

• (i)er, (i)ers, (i)ert &c ← Latin erō, eris, erit &c.
• serai, seras, sera &c ← Latin *(es)sere habeō, *(es)sere habēs, *(es)sere habet &c.
• estrai, estras, estra &c ← Latin *ess're habeō, *ess're habēs, *ess're habet &c.

Also, as we saw in 9.3.4 above, Old French had a verb ester ← stāre "to be (in a place)" with its future esterai, esteras, estra &c.

Futures from Latin erō, eris, erit &c. were found only in Old French are not attested elsewhere. If they ever occurred in Proto-Britainese they would have soon disappeared as they would have been two similar to the imperfect indicative. But there is no reason to suppose that futures similar to the other found in Old Future would also have occurred on Old Britainese. As we also observed in 9.3.4 the two verbs estr(e) and (e)star would have merged in Britainese to give a single verb "to be" as they did also in French and in Aragonese. The modern Britainese future of "to be" will be seen to be a conflation of estrai and (e)starai.

OLD BRITAINESE VARIANTS			MODERN BRITAINESE
(e)star	estr(e)	estr(e)	star
(e)starai (e)staras (e)stara(d) (e)starams (e)staradz (e)starant	serai seras sera(d) serams seradz serant	estrai estras estra(d) estrams estradz estrant	eu strá tu strá el/il strá nos strams vos stradz els/ils strá

Cf.

• Rouen and Brayon Norman: srai, sras, sra, sroms, srez, sront
• Picard: srai/sro, sros, sro, srons, srez, sront

9.6.2 Conditional Tense

As we saw above, this tense evolved from fusing the infinitive with the imperfect of habére. In practice it came to be formed by using the same stem as the future and suffixing the imperfect tense endings used by habére e.g. French je chanterais ← je chanter+(av)ais.. The tense is formed, without exception, by using the same stem as the future and suffixing the endings of the Imperfect Tense of regular verbs; see the examples below:

cantar "to sing"	baivr "to drink"	finir "to finish"	volair "to wish, want"	star "to be"
eu cantarai tu cantarai el/il cantarai nos cantaraiams vos cantaraiadz ils/els cantarai	eu bevrai tu bevrai el/il bevrai nos bevraiams vos bevraiadz ils/els bevrai	eu finirai tu finirai el/il finirai nos finiraiams vos finiraiadz ils/els finirai	eu voldrai tu voldrai el/il voldrai nos voldraiams vos voldraiadz els/ils voldrai	eu strai tu strai el/il strai nos straiams vos straiadz els/ils strai