Middle Russian Future Periphrastic Constructions in the Light of Language Contacts

ABSTRACT

This paper examines periphrastic constructions with future time reference in Middle Russian, focusing on constructions with an inchoative verb and an infinitive clause: počnu ‘begin’ + inf, načnu ‘begin’ + inf, imu ‘take > begin’ + inf, učnu ‘begin’+ inf, and stanu ‘rise > begin, become’+ inf. The author compares these periphrastic constructions and establishes that the verb učati ‘begin’ differed from the rest of the inchoatives (počati, načati, jati, stati) in that its distribution was functionally restricted and its derivational potential limited (i.e. no verbal nouns, participles, or imperfective verbs derived from this stem). On the

1 I am grateful to the two anonymous reviewers from Scando-Slavica, the editors of the volume for their valuable comments, and Dr. Katie Sykes for her professional English language editing.

basis of these features, the author argues that učnu + inf emerged due to contact influence: specifically, the borrowing of a pattern from a Finno-Ugric language spoken in the eastern or northern part of the Grand Duchy of Moscow at the time. It is likely impossible to identify the construction which served as a source for učnu + inf. The author’s hypothesis is further supported by data from contemporary dialects and early attestations in official documents from eastern and northern areas.

KEYWORDS:

• inchoative verbs
• language contact
• Middle Russian
• periphrastic future
• Russian National Corpus
• Finno-Ugric languages

0. Introduction

¹According to the well-known cross-linguistic studies conducted by Joan Bybee and Östen Dahl (Bybee and Dahl 1989; Bybee et al. 1994, 251–70; see also Ultan 1978, Bybee and Pagliuca 1987; Bybee, Pagliuca and Perkins 1991; Dahl 2000; Heine and Kuteva 2002), the most prominent sources for the development of future marking are:

• – tense-aspect forms (the perfective future in North Slavic),

• – verbs of movement (e.g. the komma future in Swedish),

• – obligation (e.g. the “have future” in Romance languages),

• – volition (e.g. Balkan “want futures”),

• – constructions with inchoative copula verbs (e.g. the werden future in German).

There are certain other less common sources which have also yielded the future, including verbs of contact such as ‘take’ (e.g. in Ukrainian and Hungarian). The existence of a variety of grammaticalization sources leads to the coexistence of, and eventual competition between, future forms which emerged from different sources within a language, i.e. to the emergence of new layers: “Within a broad functional domain, new layers are continually emerging. As this happens, the older layers are not necessarily discarded but may remain to coexist with and interact with the newer layers” (Hopper 1991, 22). The Russian language is no exception. Moreover, Middle Russian features a remarkably rich system of different periphrastic constructions with future time reference, within which constructions with auxiliaries meaning ‘begin’ play a special role.

Dahl (2000, 310) proposed a major criterion for establishing whether a language possesses a grammaticalized future tense: “future time reference obligatorily marked in prediction-based sentences.” The same grounds for defining the future tense are used in the influential work by Bybee et al. (1994, 244).

According to Dahl (2000, 325–26), Old East Slavic, together with some other North European languages, belonged to the so-called futureless area: in other words, it lacked a grammaticalized future and employed the present for future time reference. This futureless area included Germanic and Finno-Ugric (Dahl 2000, 325–26). Certain scholars argue that this is the case for Modern Russian too (see Swan 2012), but it undoubtedly holds true for Old East Slavic before the emergence of budu² + inf at the very end of the Middle Russian period. However, before the rise of budu + inf, there already existed constructions with the auxiliaries učnu ‘begin’ and stanu ‘stand, rise > begin, become’. The periphrasis with imu ‘take’ had also developed earlier, in late Old East Slavic. All these periphrastic constructions competed for the role of imperfective future, but were still under-grammaticalized; in other words, Middle Russian did not feature any analytical future form that had necessarily to be used for conveying prediction. I will further consider periphrastic constructions which could serve to refer to the future and I will follow some contact-induced patterns in their development in the “pre-future times” in the history of Russian. Thus, all the constructions I discuss will be referred to as “periphrastic futures” purely for reasons of brevity and convenience.

Our list of periphrastic constructions with future time reference in Middle Russian is completed by constructions with the auxiliaries choču ‘want’, imam′ ‘have’, načnu ‘begin’, and počnu ‘begin’, which already existed in the most archaic texts and remained in use until the end of the Middle Russian period.³

Areal and contact factors can play a significant role in the development of future tenses. Among other tenses, the future is the most likely to be affected by contact influence (Matras 2007, 46). Common patterns in the development of future markers can be found in the Balkans and Northern Europe, including the Circum-Baltic area (some generalizations can be found in Wiemer and Hansen 2012). The modern Russian imperfective future budu + inf is undoubtedly of Polish-Ruthenian origin, judging from the Middle Russian sources examined by Pennington (1968) and Moser (1998) (pace Křížková (1960, 179), who argues for independent development). Corpus-based research conducted by Penkova (2019a) has confirmed these scholars’ conclusions. The first examples of the construction budu + inf in Russian texts date back to the 15^th century; however, these constructions are limited to texts with Ruthenian linguistic features (Moser 1998). The periphrasis budu + inf made its way into Russian texts only in the 17^th century. The details of this expansion are unclear and call for separate research (on the problem of German influence on the development of the imperfective future in West Slavic, see Wiemer and Hansen 2012, 104–108).

Another periphrastic future form with the auxiliary imam′ ‘have’, lost in Modern Russian but well attested in the previous period (throughout the entirety of the Old East Slavic and Middle Russian era), also was borrowed from Church Slavonic. In vernacular Old East Slavic, as opposed to the literary Church Slavonic of the Russian redaction, there is no trace of the use of imam’ + inf (see Jur′eva 2009).

Though many studies tackle this subject matter, it remains somewhat underexplored, particularly in the light of language contacts. The present paper examines the network of Middle Russian constructions with future time reference from the perspective of contact-driven influence and with a particular focus on periphrastic futures with inchoatives as auxiliaries (imu ‘take > begin’, počnu ‘begin’, načnu ‘begin’, učnu ‘begin’, and stanu ‘rise > begin, become’).

Data and methods

The data for the study was collected from the Russian National Corpus, specifically the Middle Russian subcorpus (http://www.ruscorpora.ru/new/search-mid_rus.html). This is a highly reliable resource containing more than eight million tokens from 15^th–17^th-century sources. The corpus comprises various genres of medieval writing, from the standard and “hybrid” Church Slavonic of the Russian redaction (see Živov 2017, 205–314) to official documents and vernacular sources. Since July 2020, it has featured grammatical annotation, but it lacks lexical tagging. To extract the set of examples including periphrastic constructions from the corpus, the following masks were used: imu*/ imet*/ imeš*/ imem* for the construction imu + inf, učn* / uč′n*/ učal* / učel* for the construction učati + inf, stanu* / stane* / stal* for the construction stati + inf, budu* / bude* for the construction budu + inf, počn* / poč′n / počal* for the construction počati + inf, načnu*/ nač′nu* / načne*/ nač′ne* / nača for the construction načati + inf, and imam* / imaš* / imat* / imut* for the construction imam′ + inf. These masks cannot always give fully disambiguated results for the purposes of this research; all data driven from the corpus was then selected manually.

The final dataset contains 3448 examples of periphrastic constructions with future time reference, i.e. all occurrences of the following constructions: učnu + inf, imu + inf, počnu + inf, načnu + inf, imam′ + inf, stanu + inf, and budu + inf.

This paper will begin with a brief overview of the common types of periphrastic constructions known to exist in the Finno-Ugric languages. This information will serve as the background to a discussion of Middle Russian periphrastic constructions. The following sections will survey periphrastic constructions with future time reference in Old East Slavic and Middle Russian. In the fourth section, I will discuss the distribution of the auxiliaries that emerged out of inchoative verbs in Middle Russian, focusing on differences in their token and clause-type frequency and syntactic behavior. This section will be based on the results of the corpus research conducted by Penkova (2019a). In the last section, I will compare the inchoative constructions and suggest an explanation for the contact-driven emergence of one of them.

1. FTR⁴ periphrastic constructions in Finno-Ugric

This overview is based on a study by Helle Metslang (1996). She provides a classification of periphrastic constructions with future time reference in the Finno-Ugric languages, taking into account their grammaticalization sources. The present paper will consider only items of her classification that are essential for discussing the situation in Middle Russian. In the following table, taken from Metslang 1996, Finno-Ugric periphrastic constructions are classified according to their general meaning, their so-called grammaticalization source (or family; here, the original meaning of the respective auxiliary), and the Finno-Ugric languages in which they occur.

Table 1. FTR periphrastic constructions in Finno-Ugric

General meaning	Auxiliary	Finno-Ugric languages
Agent-oriented modality	SHALL	Sami languages, Hungarian, Finnish, Old Standard Finnish, Old Standard Estonian
Agent-oriented modality	WILL / WANT	Sami languages, Finnish, Old Standard Finnish, Old Standard Estonian, Old Standard Hungarian
Movement	COME	Finnish
Contact^5	TAKE, SEIZE > BEGIN	Estonian, Livonian, Votic, Hungarian, Mansi
Inchoativity	BEGIN	Occurs in all branches
Inchoativity	BECOME	Finnic: Livonian, Karelian, Estonian, Finnish dialects
Change of position	RISE > BECOME	Votic, Izhorian
BE-copula	BE (LE-stem)	Finnic, Mari, Permic languages, Hungarian

According to Metslang, COPULA types emerged due to convergence with Slavic, while the SHALL, WILL, and COME types emerged under contact with other European languages spoken outside indigenous Finno-Ugric territories.

This is not the case for the BEGIN type. According to Metslang, the BEGIN type is the most widespread source for the future tense in all branches of Finno-Ugric. Moreover, Finno-Ugric languages tend to develop inchoative auxiliaries from other semantic sources:

The constructions, occurring most often in all branches of Finno-Ugric languages but almost nowhere else are the BEGIN-type. Already Proto-Finno-Ugric (or even Proto-Uralic) seems to have had some elements of the future – polysemy of ingressivity and futurity. Finno-Ugric languages develop their futures from whatever ingressive constructions. (Metslang 1996, 138)

For example, the Estonian verb hakkama, the Livonian akkõ, and the Hungarian fog, operating as future auxiliaries, originally denoted seizing or taking and developed into futures via an ingressive stage (Metslang 1996, 138). There are also many BEGIN futures in the Finnic and Volgaic languages: the Karelian verb rubia (e.g. rubiav pidämäh ‘it will be needed’), the ingressive verb ajke- used for FTR in Skolt Sami, the Mordvinian verb karmams ‘begin’ (e.g. Erzya (Mordvinic or Mordvin) karmat lovnomo ‘you will read’), etc. (Metslang 1996, 138, Dahl 2000b, 324).

Relying on the conclusion reached by Metslang, I will further argue that (at least) one BEGIN future in Middle Russian emerged under the areal contact-driven influence of the Finno-Ugric languages. The arguments for such a development will be set out at the end of this paper after more detailed discussion of the network of FTR periphrastic constructions in Middle Russian.

2. Old East Slavic periphrastic constructions with future time reference

The initial Old East Slavic⁶ system reflected in the most archaic sources was futureless; there was no grammaticalized future, either synthetic or periphrastic, although there were ways to convey future meaning, such as choču (WANT) + inf, imam′ (HAVE) + inf, počnu (BEGIN₁) + inf and načnu (BEGIN₂) + inf⁷.

Table 2. FTR periphrastic constructions⁸ in Old East Slavic

General meaning	Auxiliary meaning	Construction	Distribution
Possession	HAVE	imam’ + inf	Church Slavonic
Agent-oriented modality	WANT	chošču + inf choču + inf,	Church Slavonic, Old East Slavic
Inchoativity	BEGIN	počnu + inf	Old East Slavic
Inchoativity	BEGIN	načnu + inf	Church Slavonic, Old East Slavic
Contact	TAKE, SEIZE > BEGIN	imu + inf	Old East Slavic

However, the construction with WANT mostly expressed volition and intention, and only marginally the prospective meaning (see Jur′eva 2009, 97–143), while the auxiliary could be used in the present form as well as the preterit form (with avertive meaning, as in (2)). The HAVE type was borrowed from Church Slavonic and was characteristic of high-code texts only. It was used mostly with the meaning of necessity (as in (1); see Jur′eva 2009, 72). The WANT and HAVE auxiliaries could combine with both imperfective and perfective infinitives:

(1)	i ne imašihave.AUX.PRS.2SG v′′zvratitisjareturn.INF.PFV k materi svoei ‘you will never return to your mother’ (Alexandria,9 an Old East Slavic translation from Greek, not later than the 13th century)
(2)	i chotěwant.AUX.AOR.3SG letětifall.INF.IPF, i tu podchytiša i pod rucě ‘he was about to fall, and they then grabbed him by the arms’ (Kievan Chronicle, compiled in 1199, f. 166v)

The verb načat′ in modern Russian has restrictions on its compatibility with infinitives: it cannot combine with a predicate denoting a state or a constant attitude (*načnu imet′ ‘begin to have’). As shown by Jur′eva (2010), in Old East Slavic, the verbs počati and načati ‘begin’ did not have such restrictions on their compatibility. Constructions with počati and načati could be used to designate not only the initial phase of a process (for situations in which initial, middle and final phases could be clearly distinguished (3)), but also to designate the emergence of a new situation with predicates denoting processes (4) or states (5) that cannot be semantically separated into different phases, e.g.:

(3)	Priim jabloko i isprosi nož’ da po obyčaju rěža e načnet′begin.AUX.PRS.3SG gryzticrunch.INF ‘taking an apple, he asked for a knife so that, cutting it according to the custom, he might start eating [it]’ (History of the Jewish War, an Old East Slavic translation from Greek)
(4)	I rěša bolgare toli ne budi mira meži nami. oliže kamen′ načnet′begin.AUX.PRS.3SG pl(a)vatifloat.INF a chmel′ grjaznutisink.INF ‘and the Bulgarians said: let there be no peace between us, when the stone begins to float and the hops to sink’ (Primary Chronicle)
(5)	Daže v tom ustoit′ i o(t′)c(e)m′ mja imětihave.INF počnet′begin.AUX.PRS.3SG vo pravdu i dobra moego hotětiwish.INF to jaz i s(y)nom iměju sobě ‘if he will endure this and will consider me a father and truly wish me well, then I will consider him a son’ (Kievan Chronicle, compiled in 1199)

According to Andersen (2006b), the počnu/načnu (BEGIN_1/2) + inf type is characteristic of East Slavic, despite its scarcity in Old Church Slavonic. Unlike in the case of HAVE and WANT, the dependent infinitive clauses of this type are exclusively imperfective. Andersen (2006a) claims that in the Old Novgorod dialect there already existed a periphrastic future with the auxiliary počati (BEGIN₁). Jur′eva has shown that the BEGIN₁ and BEGIN₂ types cannot yet be considered periphrastic futures in Old East Slavic, as they were much more frequently used with the auxiliary in the past tense form than in the present tense (Jur′eva 2016, 545–46):

(6)	načabegin.AOR.3SG molčatikeep_silent.INF na mnogy časy ‘[he] began to keep silent for hours’ (Life of Andrew the Fool, a 12th-century Old East Slavic translation from Greek)

In late Old East Slavic, a new periphrastic construction with the auxiliary imu (TAKE, SEIZE) and an imperfective infinitive clause emerged:

(7)	niktože imet′take.AUX.PRS.3SG im pakostitiinterfere.INF ‘nobody will interfere with them’ (Trade agreement of Smolensk with Riga and Gotland compiled in 1229, copy D)

The earliest attestations of the periphrastic construction with imu (TAKE) date back to the mid-1200s and occur in the treaties of Smolensk with Riga and Gotland. These texts originate from the Western East Slavic area contiguous to the Circum-Baltic area. The same TAKE type is also attested in some Finnic languages, such as Estonian (“the meaning of the Estonian verb hakkama ‘begin,’ aiming at a future auxiliary was originally ‘seize’,” Metslang 1996, 127), Livonian, and Votic, e.g.:

(8)	Votic: Miä ävitin sermuhsē; tšen veta-btake.AUX-3SG leütä? ‘I lost the ring; who will find?’ (Ariste 1968, 72)

It seems unlikely that the TAKE type developed independently in Old East Slavic and the Finnic languages, since verbs of contact are cross-linguistically very infrequent as a source for the future (Heine and Kuteva 2002, 288).

The TAKE type was the first East Slavic attempt to develop an imperfective future (see Jur′eva 2009, 32), and it still exists in Ukrainian as well as in some Belarusian dialects (Jankoŭski 1983, 200). As Ševeleva (2017) has shown, the original meaning of the imu + inf constructions was not modal, as was previously thought, but inchoative (‘take something to do’ > ‘start something to do’).

Sometimes, the auxiliary imu (TAKE, SEIZE) is confused with imam' (HAVE) by reason of their phonetic similarity and shared homonymous third person plural forms. Metslang and Dahl (2000, 319) make this mistake, stating that Ukrainian possesses a future which developed from a source with the original meaning HAVE. This same attraction and confusion between the two constructions is a feature not only of the scholarly literature, but also of late Old East Slavic and early Middle Russian texts. The HAVE construction lost its modal meaning under the influence of the vernacular TAKE construction and began to be used as a high-code counterpart of the latter (see Jur′eva 2011), e.g.:

(9)

ašče li po moem živote oskudevati načnet monastyr′, černorizec′ imat′have.AUX.PRS.3SG malitisjadecrease.INF i potreby manastyr′skija načnut oskudevate [sic!], to vědušče budite, jako ne ugodil esm′ Bogu ‘If after my death the monastery becomes poorer, [the number of] monks decreases, and the needs of the monastery begin to decrease, then know that I did not please God’ (Moscow chronicle, 16th century, Jur′eva 2011, 83)

3. Middle Russian FTR periphrastic constructions

In the Middle Russian period, new periphrastic constructions emerged and started to compete with the old ones mentioned in the previous section. Below is a list of old and new FTR periphrastic constructions in Middle Russian, with new constructions in bold:

Table 3. FTR periphrastic constructions¹⁰ in Middle Russian

General meaning	Auxiliary meaning	Construction
Possession	HAVE	imam′ + inf
Agent-oriented modality	WANT	choču + inf
Contact	TAKE, SEIZE > BEGIN	imu + inf
Inchoativity	BEGIN1	počnu + inf
Inchoativity	BEGIN2	načnu + inf
Inchoativity	BEGIN3	učnu + inf
Change of position	RISE > BEGIN, BECOME	stanu + inf
BE-copula	BE	budu + inf

The first construction, which appeared in the 15th century, employed the auxiliary u-čn-u ‘begin’ with the same root as in po-čn-u (BEGIN₁) and na-čn-u (BEGIN₂), e.g.:

(10)

A čerez sju moju gramotu kto učnetbegin3.PRS.3SG vstupatis′interfere.INF ili čem obiditioffend.INF, byti ot menja, ot velikogo knjazja, v kazni ‘And if anybody interferes or offends in contravention of the present act, he shall be punished by me, the Grand Prince’ (Charter from Grand Prince Vasilij Vasil'evič to Metropolitan Iona forbidding the prince’s officials, fishermen, and hunters to enter the Vyksinga River of the Metropolitan Beloozerski Čerepovec Monastery, 1448–1461)

However, there are some significant differences between učnu and the old BEGIN auxiliaries discussed below.

Comparing modern Russian načat′ ‘begin’ and stat′ ‘become,’ Korotkova and Saj (2006, 96–97) conclude that the former focuses on the phase structure of P, while for the latter, the phase structure of the event is irrelevant, and attention is focused on the contrast between the moment when P (proposition) occurs and the time preceding it, when P was not taking place. This was not the case for the auxiliary učati (cf. above, the same comment on the semantics of načati and počati), which demonstrates semantic features similar to stat’ and not to načat’ in that it focuses on the emergence of a new situation. This explains its ability to combine with predicates denoting states and events that cannot be divided into phases, e.g.:

(11)	A tě dvatcat′ stolbov zdělat′ s peremyčkoju, protiv prežnevo obrasca, ili kak učnetBEGIN3.AUX.PRS.3SG ukazyvat′SHOW.INF podmasterja Dmitrej Kostousov ‘and those twenty columns should be made with a lintel, in accordance with the old pattern, or as the journeyman Dmitrij Kostousov will show’ (Income and Expense Books of the Privy Prikaz, 1665–1666)
(12)	Egda že sei zakon povsjudu vsjak uslyšit, togda … vsjakaja žena načnet muža čtiti i slovesem ego ne učnetBEGIN3.AUX.PRS.3SG protivnaOPPOSED.ADJ by[ti]BE.INF ‘When everyone hears this law everywhere, then … every wife will begin to honor her husband and will not object to his words’ (Play of Artaxerxes, 1672)

The earliest attestations of the RISE > BEGIN, BECOME type with the auxiliary stanu (RISE > BEGIN, BECOME¹¹) date back to the very end of the 15^th century. This auxiliary is always regarded as belonging to the BECOME type (Korotkova and Saj 2006), but its grammaticalization clearly developed out of its “change of position” and locative uses with an infinitive expressing purpose (see Penkova 2021).

Moldovan (2010) discusses stylistic differences between the periphrastic constructions in question in Middle Russian. According to the scholar, imam′, choču, and načnu auxiliaries are typical for standard Church Slavonic texts until the 16^th–17^th centuries, while počnu prevails in hybrid Church Slavonic and in vernacular texts. However, Moldovan does not specify the place of the učnu + inf construction in the network of Middle Russian constructions.

The last periphrastic construction to emerge in Middle Russian was budu (BE) + inf, which is supposed to be a loan from Polish via Ruthenian (see above). Unlike učnu and stanu, budu did not have corresponding past tense forms that could combine with an infinitive clause. This construction eventually won the competition for the role of analytical imperfective future. For a discussion of why the budu type forced out all the other periphrastic constructions, see Swan 2012. Stojnova has also made a significant contribution to research into the competition between stanu and budu, but on the basis of Modern Russian material alone (Stojnova 2019). The competition between inchoative futures in Middle Russian is studied in detail in Penkova 2019a. I will give an overview of the results and conclusions from this study in the next section before returning to the probable contact effects at play.

4. Distribution of auxiliaries that developed out of inchoative verbs in Middle Russian

At first glance, it seems almost impossible to pinpoint any real difference in the distribution of the inchoative periphrastic constructions in question. They all seem to produce a similar meaning and can occur in the same syntactic context. For instance, imu and počnu occur in the same context in (13), as do stanu and učnu in (14):

(13)

A imuttake.AUX.PRS.3PL nas svaživatipit.INF tatarove, počnutbegin1.AUX.PRS.3PL mi davat[i]give.INF velikoje knjaž[e]nije, i mně ne vzjat[i] ‘And if the Tatars pit us against each other, and offer me the Grand Princedom, I should not accept’ (Agreement of Grand Prince Vasilij Vasil′evič with Prince Vasilij Jur′evič, 1439)

(14)

A prodadut tu sol′ na Voloce i po torgom i vo Rževe i v Degunine, a čto stanutrise.aux.prs.3pl kupit′buy.INF ne na monastyrskuju potrebu da učnutbegin3.AUX.PRS.3PL to opjat′ prodovat′sell.INF, čto kupjat, – i naši tomožniki i pošlinniki s togo u nich tovaru tamgu i pošlinu emljut ‘And if they sell that salt in Volok and in the markets in Ržev and Degunin, and if they buy not for the monastic need and then sell again what they buy, then our customs officers and collectors of duties on that product collect tax and duty from them’ (Letter of Preference granted by Grand Prince Vasilij III Ivanovič to Nifont, abbot of the Iosif-Volocky monastery, on duty-free travel to Novgorod, Ruza, and Beloozero, October 20, 1527)

Although the inchoative verbs počnu (BEGIN₁), načnu (BEGIN₂), and učnu (BEGIN₃) share the same root, periphrastic constructions employing these auxiliaries have remarkably different distributions. According to the corpus study conducted by Penkova (2019a), the inchoative “futures” in question have different token frequency in their past (PST or AOR) and present tense (PRS) forms. The Middle Russian Corpus gives the following picture:¹²

Table 4. Distribution of constructions with inchoative auxiliaries in past and present tense forms (according to the Middle Russian corpus¹⁴)

Auxiliary	XV^th century	Ipm^13	XVI^th century	ipm	XVII^th century	ipm
počatiPRS + inf	23	23	21	6	9	2
počatiPST + inf	105	108	167	48	172	47
načatiPRS + inf	22	22	66	19	60	16
načatiAOR.3SG + inf	333	901	872	255	271	74
statiPRS + inf	0	0	57	16	221	61
statiPST + inf	4	5	42	12	178	49
učatiPRS + inf	21	21	964	282	1148	317
učatiPST + inf	4	5	135	39	158	43

At the initial stages of development, one of the most significant indicators of grammaticalization is the high frequency of a construction. As is evident from the table, the frequencies of stati (RISE > BEGIN, BECOME) in the past and present are roughly equal, whereas the present forms of učati (BEGIN₃) predominate. At the same time, počati (BEGIN₁) and načati (BEGIN₂) are largely used in their past tense forms. Evidently, the old BEGIN constructions were far from having grammaticalized as periphrastic futures, unlike the new BEGIN₃ construction, the primary function of which was to refer to the future.

In this section, attention will be paid to the constructions with imu, učnu, počnu, načnu, stanu, and budu, as well as to the imam′ (HAVE) construction, as the imu (TAKE) and imam’ (HAVE) types share homonymous third person plural forms. The homonymous forms are treated separately as the data in the corpus is not disambiguated.

As is evident from table 5 (data from Penkova 2019a are supplemented with data on the frequencies of the constructions počnu + inf and načnu + inf), stanu and budu constructions had a very low token frequency in the Middle Russian period. By contrast, učnu demonstrates an extremely high token frequency in the 16^th and 17^th centuries. Thus, the 16^th century is a crucial period for the development of the periphrastic future in Middle Russian. In the 16^th century, učnu + inf became the most frequent periphrastic construction with a future time reference: its frequency skyrockets, while the frequency of imu falls abruptly. Such simultaneous changes can hardly be considered accidental. According to Penkova 2019a, učnu and imu constructions, unlike other “periphrastic futures,” also shared similar clause-type frequencies¹⁵ for their occurrence in different clauses: both were very likely to appear in conditional and relative clauses and very unlikely to appear in other clauses. Both imu and učnu futures should be regarded as non-assertive futures used for background and potential contexts (mostly conditional and correlative clauses, see Penkova 2019a for details), e.g.:

(15)	сhto … imettake.AUX.PRS.3SG žitilive.INF našich bojar v tvoei vočine, bljusti ich, kak i svoich ‘whichever of our boyars will live in your fiefdom, you should treat them in the same way as yours’ (Agreement of Grand Prince Vasilij Dmitrievič with the princes of Možajsk and Dmitrov Andrej Dmitrievič and Petr Dmitrievič, 1401–1402)
(16)	chto v tech pustošach i v derevnjach učnetbegin3.AUX.PRS.3SG žitilive.INF ljudei, i ne nadobe im tjanuti k stanom ničem ‘whichever of the people will live in those wastelands and the villages, they need not pay taxes to the administrative districts’ (Charter of Exemption from Grand Prince Vasilij Dmitrievič to the village of Vjachorev and the wasteland, 1417–1425)

The identical distribution of učnu and imu constructions allowed Penkova (2019a) to conclude that učnu was a successor of the imu construction and substituted for it not only chronologically (see Table 5), but also functionally.

Table 5. Token frequency of constructions with auxiliaries in the present tense forms (according to the Middle Russian corpus)

Auxiliary	XV	%	XVI	%	XVII	%
imu + inf	152	40%	20	1.4%	1	0.06%
imutPRS.3PL + inf	48	12.6%	41	3%	43	2.56%
imam’ + inf	114	30%	204	14.7%	106	6.31%
stanu + inf	0	0%	57	4.1%	221	13.2%
počnu + inf	23	6.1%	21	1.5%	9	0.53%
načnu + inf	22	5.8%	66	4.7%	60	3.57%
učnu + inf	21	5.5%	964	69.4%	1148	68.41%
budu + inf	0	0%	17	1.2%	90	5.36%
Total FTR	380	100%	1390	100%	1678	100%

5. Explanations for the areal-contact emergence of the učnu construction

In this section, I will argue that učnu + inf emerged in Middle Russian as a replication of a Finno-Ugric model with an inchoative verb or other marker of inchoative meaning. Arguments for the areal-contact emergence of učnu + inf fall into three groups: typological arguments, functional arguments, and source evidence.

5.1. Typological arguments

As was pointed out in section one, BEGIN as a grammaticalization source for marking future-time reference was a typical pattern in Finno-Ugric, and should be regarded as a characteristic feature of the Finno-Ugric languages. According to the study of Thomason and Kaufman (1988, 238–51), among different possible types of contacts between East Slavic and Finno-Ugric, a language shift was one of the prevailing. A shift-induced situation involves interaction between languages in which a group of speakers of one language switches to a second language when communicating with each other, depending on various factors. The “mistakes” made by a group of speakers speaking a new language replicate the structures of their original language. According to a recent study by Sakel (2007), there are two main types of borrowing: borrowings of matter, and pattern borrowing. The former refers to the borrowing of a language item as a whole; the latter refers to “the case where only the patterns of the other language are replicated, i.e., the organization, distribution, and mapping of grammatical or semantic meaning, while the form is not borrowed” (Sakel 2007, 15). In the context of interference through a language shift, pattern borrowing is more typical. The replication of a Finno-Ugric BEGIN construction (or inchoative marker) as a mere substitution for the imu construction would therefore have been likely.

5.2. Functional arguments

Functional arguments fall into two types, concerning either the restricted distribution of the inchoative učati as compared to the other inchoatives počati and načati, or the absence of traces of grammaticalization. In the following section, I examine these criteria more closely, starting with the former.

5.2.1. The criterion of restricted distribution

Section four discussed differences in token frequency between various periphrastic constructions with inchoatives. It was pointed out that učati had a different distribution to the other inchoatives with the same root č′n-‘begin’, instead patterning with imu. There are also certain other features which distinguish učnu not only from počnu and načnu, but also from stanu and even from imu. The use of the verb učati was restricted to infinitive clauses alone, and it never occurred as a matrix non-auxiliary verb, unlike all the other inchoatives, which also functioned as matrix verbs, e.g.:

(17)	Medljaja počinaibegin1.IMP.2SG dělo, počenbegin1.PTCP že v′′borzě končai ‘Start an affair slowly, but having started, finish it fast’ (Pčela, an Old East Slavic translation from Greek produced no later than the 13th century)
(18)	Vas bo istjažet Bog, ašče s lenostiju načnetebegin2.PRS.2PL delowork.ACC.SG božije ‘God will punish you if you begin God’s work with laziness’ (Theodosius Byvalсev, bishop of Rostov and Jaroslavl′. Formal Summary of the Presbyter’s Teaching, 1454)
(19)	Stanemstand.PRS.1PL u Donu protivu knjazja Dmitria, dokolě prislějet′ k nam′ s′′větnik′ naš′ Yagajlo s svoeju siloju ‘We will stand by the Don against Prince Dmitry until our ally Jagiello sends [help] to us with his strength’ (Chronicles of the Battle of Kulikovo, 1380–1400)

Furthermore, učnu, unlike all the other inchoative auxiliaries, had very restricted derivational potential. Neither participles nor verbal nouns could derive from this verbal stem in Middle Russian. Other inchoatives do not share such derivational restrictions, сf.: počati_INF – počatie_VN; načati_INF – načalo_VN, načatie_VN; jati_INF – njatie_VN; stati_INF – statie_VN, učati_INF – *učatie_VN.

Finally, učati lacks an imperfective counterpart in Middle Russian, unlike the other inchoatives, сf.: počati_PFV – počinati_IPF, načati_PFV – načinati_IPF, jati_PFV – imati_IPF, stati_PFV – stavati_IPF and stojati_IPF, učati_PFV – *učinati_IPF.¹⁶

5.2.2. The criterion of lacking traces of development

There is no evidence for any particular construction that might have served as a grammaticalization source for učnu + inf. However, specific source constructions can be identified for other inchoative “futures.” The construction stanu + inf probably emerged from the use of stanu denoting ‘stand or stay’ and a purpose infinitive (or supine) clause (Penkova 2021):

(20)	I pride s˜toslav k porogom. I ne bě lzě proiti porogov. I stastand/remain.AOR.3SG zimovat′winter.SUP v běloberež′ii ‘And Svjatoslav came to the riffles, and it was impossible to pass the riffles, and he remained to winter in Beloberež′e’ (Primary Chronicle)
(21)	stalstand_up.PST.3SG pred obrazom prečistyje Vladyčicy Bogorodicy molitisjapray.INF so slezami ‘[he] stood up before the image of our Holy Lady to pray with tears’ (Ivan Peresvetov. The Great Petition, 1549)

We can also suggest a probable grammaticalization source for the construction imu + inf. The use of the matrix verb imu, which could co-occur with noun phrases, could easily have expanded to infinitive clauses (imu + NP > imu + inf), e.g.:

(22)	Ašče ne imetecapture.PRS.2PL volchvu seju i ne idu ot vas ‘If you do not capture these two magi, then I will not leave you’ (Primary Chronicle)

We cannot trace particular source constructions for počnu and načnu + inf since these periphrases occur even in the most archaic texts, having already undergone grammaticalization in the prehistoric period. That said, this process involves the same extension from noun phrase to infinitive clause as was suggested for imu (počnu + NP > počnu + inf).

Table 6 summarizes the distinctive features of inchoatives according to the criteria mentioned above.

Table 6. Features of inchoatives

Type	Verb	Non-auxiliary use	Verbal noun	Participle	Imperfective	Source construction
BEGIN1	počnu	+	+ počatie	+ počav- / počen-	+ počinati	+
BEGIN2	načnu	+	+ načalo, načatie	+ načav-/načen-	+ načinati	+
BEGIN3	učnu	–	–	–	–	–
TAKE, SEIZE	imu	+	+ njatie	+ im-	+ imati	+
RISE > BECOME	stanu	+	+ statie	+ stav-	+ stavati, stojati	+

The lack of features for učnu presented in Table 6 confirm the hypothesis that učnu was alien to the Middle Russian verbal system and that the construction učnu + inf could have entered the system of periphrastic “futures” as the result of external influence. The best candidates for such influence would have been the now extinct Finno-Ugric languages spoken in the central and eastern parts of the Grand Duchy of Moscow, such as Merja, Meščerian, Muromian, or other languages spoken to the north, which, like the other Finno-Ugric languages, most likely had the begin future or at least an inchoative marker at their disposal (cf. also Veps, or Vepsian language, featuring an inchoative suffix -škande-). Some comprehensive observations on the effect of contact configuration for Slavic morphosyntax in general and the effect of Finno-Ugric interference in Russian morphosyntax, in particular, can be found in Thomason and Kaufman (1988) and earlier studies cited by them, cf. also Seržant (2021). Archaeological evidence of close contacts between the Eastern Slavs and the Finno-Ugric tribes of the Volga-Oka basin in the Middle Ages is discussed in Rjabinin 1990.

5.3. Source arguments

Since typologically and functionally učnu + inf meets the criteria for a borrowed item, it is worth examining the areal-dialectal attribution of the sources in which the construction in question occurs. These sources fall into two categories: the sources of the earliest attestations and the sources of relic attestations.

5.3.1. Sources of the earliest attestations

Data from the Russian National Corpus suggest that the earliest attestations of učnu + inf appeared in official writing originating from the east-central and northern areas of the Grand Duchy of Moscow, e.g.:

(23)	Moscow: I сhto v teсh pustošaсh i v derevnjaсh učnetbegin3.PRS.3SG žitilive.INF ljudei, i ne nadobe im tjanuti k stanom ničem ‘And whichever of the people will begin to live in the wastelands and the villages, they need not pay any duties to the administrative districts’ (Charter of Exemption from Grand Prince Vasilij Dmitrievič to the village of Vjaсhoreva and the wasteland, 1417–1425)
(24)	Beloozero: I chto v tech počinkach učnetbegin3.PRS.3SG žitilive.INF ljudei, i tem ljudem ne nadobe ich knjažaja dan′, ni pisčaja belka, ni kotoraja inaja dannaja pošlina ‘And whichever of the people will begin to live in those new villages, those people need not pay the princely tribute, nor the fee for the census of lands, nor any other fee’ (Charter of Legal and Financial Exemption from Michail Andreevič, prince Verejsko-Belozerskij, to the abbot of the Rostov Borisoglebskij monastery on land along the rivers Kovže, Kem and others in Beloozero, 1432–1486)
(25)	Čerepovec: A čerez sju moju gramotu kto učnetbegin3.PRS.3SG vstupatis′intervene.INF ili čem obiditioffend.INF, byti ot menja, ot velikogo knjazya, v kazni ‘And if anybody, not taking into account this document, begins to intervene or to offend in any way, he shall be punished by me, the Grand Prince’ (Charter from Grand Prince Vasilij Vasil'evič to Metropolitan Iona forbidding the prince's officials, fishermen, and hunters to enter the Vyksinga River of the Metropolitan Beloozerskij Čerepovec Monastery, 1448–1461)
(26)	Tver′: [i] kto učnetbegin3.PRS.3SG žitilive.INF v tom dvore ljudej, ino tem ljudem ne nadobe ni dan′, ni jam ‘[and] whichever of the people begins to live in that unit of taxation, those people need pay neither a tribute nor a duty for transportation’ (Charter of Michail Borisovič, Grand Prince of Tver’, to Vassian, abbot of the Trinity-Sergius Monastery, on the courtyard in Kašin, 1461–1466)
(27)	Archangelsk: A сhto učnetbegin3.PRS.3SG ot′′imatitake_away.INF u svjatago Nikoli to selo, sužjusja s nim pered Bogom ‘And whosoever will begin to take away the village from the monastery of St. Nicholas, I will stand trial with him before God’ (Donation of Osip to the Nikolaev Čuchčenemskij monastery in the village of Stanšinskoe, which is on the Lukin coast, 1450–1500)

One might argue that the frequency of učnu + inf in Middle Russian official writing throughout the 16th century (see Tables 4 and 5 above) militates against its contact emergence. However, the further expansion of učnu + inf can easily be explained by the prestige of the dialect of Moscow and the adoption of official Muscovite writing traditions in other parts of the Grand Duchy of Moscow.¹⁷

5.3.2. Relic attestations

Relic attestations of učnu + inf in contemporary Russian dialects are largely limited to the north-eastern areas of the former Grand Duchy of Moscow, just as the earliest evidence from the Middle Russian corpus was. Some examples can be found in the northern dialects of the Archangelsk and Vologda regions, with others in the east-central dialects (specifically the dialect of Jaroslavl′, which borders with the Vologda region), e.g.:

(28)	Archangelsk: Gorazd zubarit′-to, a puščaj drugoj nad im učnetbegin3.PRS.3SG zubarit′taunt.INF, tak nebos′ drat′sja polezet ‘[He is] skilled at taunting, and let the other begin to taunt him, he will likely get into a fight’ (Podvysockij 1885, 57)
(29)	Vologda: On kak učnetbegin3.PRS.3SG šti chlebat′slurp.INF, dak i ne ostanovit′ ‘Once he begins to slurp soup, it is impossible to stop him’ (Panikarovskaja 2005, 163)
(30)	Vologda: Kak učnubegin3.PRS.1SG buzgat′beat.INF, tak menja uznaeš′ ‘As I start to beat, you will recognize me’ (Levičkin and Myznikov 2006, 39)
(31)	Jaroslavl′: Budeš′ nebritoj valjatsja v izbe/ Čёrtiki prygat′jump.INF učnutbegin3.PRS.3PL po tebe ‘If you lie unshaven in a hut, devils will start jumping on you’ (Mel′ničenko 1991, 10, 25).

Outside this area, attestations of učnu + inf are found only in the Russian dialects of the Republic of Mordovia, where some Finno-Ugric languages such as Erzya and Mokša (both featuring periphrastic begin-future) are still spoken, cf.:

(32)	Republic of Mordovia: Učnubegin3.PRS.1SG k ja kapat′dig.INF usat p′′manen′ku ‘Let me start digging / I will start digging the yard slowly’ (Semenkova 2006, 8, 115)
(33)	Mokša: afNEG karma-nbegin.AUX.FUT-1.SG jaka-mawalk-INF ‘I will not be walking’ (Fournet 2012, 4)

Conclusions

The analysis carried out in this paper supports the hypothesis that the auxiliary učati ‘begin₃’, which differed significantly from all other inchoatives occurring with infinitives, entered the system of inchoative auxiliaries as the result of external influence, replicating a Finno-Ugric model with an inchoative verb or an inchoative marker (reinterpretation or remodeling also is not excluded, cf. a discussion of this phenomenon in Thomason and Kaufman 1988, 61–62). The construction with učati ‘begin₃’ meets the relevant criteria of restricted distribution and a lack of source constructions. Our hypothesis proposing replication of a Finno-Ugric model fits with the type of contact, namely interference through a language shift, prevailing between Middle Russian and the Finno-Ugric languages spoken in the Grand Duchy of Moscow. Furthermore, the expansion of constructions with inchoative auxiliaries in Old East Slavic and Middle Russian (e.g. the development of inchoatives from the verb of contact imu ‘take’ and from the change of position verb stati ‘rise’, mirrored by similar processes in Finno-Ugric) could have occurred due to contact influence from contiguous Finno-Ugric languages.

The proposed hypothesis is barely the first attempt to tackle the emergence of the construction učnu + inf, which requires further exploration. There are still many questions remaining without answers. Foremost, it is necessary to identify the exact source-language and the specific marker that served as a pattern for the replication.

Notes

1 I am grateful to the two anonymous reviewers from Scando-Slavica, the editors of the volume for their valuable comments, and Dr. Katie Sykes for her professional English language editing.

2 Here and below, the first-person singular form is used as an abbreviation of the respective auxiliary used in a periphrastic construction.

3 For the future anterior, or the so-called “second compound future” in Old and Middle Russian, rarely used for future time reference, see Penkova (2019a; 2014; 2019b).

4 Future time reference

5 The notion of contact incorporates not only states, but also dynamic actions.

6 Old East Slavic period covers the period from the 11th until the 14th centuries, Middle Russian — from the 15th until the 17th centuries.

7 There is also another inchoative verb usually mentioned in Old Church Slavonic grammar – v’’čati, but it only marginally occurs in Old Church Slavonic as well as in Old East Slavic sources (cf. SRJa 11–17, 170) and, thus, cannot be considered an auxiliary for periphrastic constructions.

8 Constructions with FTR are classified according to their general meaning and the original meaning of the auxiliary in question.

9 Here and onwards, citations of Old East Slavic and Middle Russian sources are extracted from the Russian National Corpus (NKRJA).

10 Constructions with FTR are classified according to their general meaning and the original meaning of the auxiliary in question.

11 The attribution of a particular auxiliary to a certain family gram may change over time. For this reason, we refer to such changes by introducing a particular evolutionary path, i.e. RISE > BEGIN, BECOME; TAKE, SEIZE > BEGIN, etc.

12 Inchoatives should be considered according to their semantics, distinguishing between uses with a focus on the initial phase of a situation and uses which focus on the emergence of a new situation. However, most contexts are ambiguous, so at this stage, it is impossible to estimate the frequency of each of these uses. Moreover, the Middle Russian corpus is not lexically annotated, so frequency can be estimated only in instances per million words.

13 Ipm – Instances per million words.

14 The whole Middle Russian corpus contains more than eight million tokens: 1010714 word forms in the 15th-century text corpus, 3452211 word forms in the 16th-century text corpus, and 3613710 word forms in the 17th-century text corpus at the time of accessing.

15 Joan Bybee and Sandra Thompson (1997, 378) propose distinguishing two types of frequency: token frequency and type frequency. The latter refers to the number of different lexical items a construction is applicable to. Studying this phenomenon did not give significant results. However, comparing frequencies of periphrastic constructions in different types of clauses, i.e. measuring clause-type frequency, turned out to be much more revealing (Penkova 2019a).

16 Imperfective counterpart is not attested in Middle Russian but is known in some modern Russian dialects.

17 The problem of the further disappearance of učati in Standard Russian is not connected with language contacts and calls for separate research.