How the cause and purpose syncretism was lost: the tale of two Russian subordinators

1 Setting the stage

In nineteenth-century Russian, complex subordinators with dlja togo (lit. ‘for it’) and a simplex complementizer could convey both causal meanings, as in (1), and purpose meanings, as in (2).^[1]

The example in (1) clearly conveys the causal meaning: the main clause describes a non-agentive situation, while the indicative subordinate clause presents a state of affairs assumed to be true and independent. In contrast, (2) expresses purpose: the subordinate clause does not assert or assume truth but refers to a potential future situation the agents of the main clause view as desirable when acting.

The starting point of my analysis is that, over time, causal uses of dlja togo became obsolete: (1) is no longer possible in contemporary Russian, while (2) remains grammatical. A similar, though less drastic, shift occurred with subordinators based on zatem (lit. ‘behind it’).

My goal is to examine the decline of the causal uses of these subordinators from a corpus-based micro-diachronic perspective. This is a challenge, as both typology and Russian linguistics have focused more on the emergence of causal clauses than on their decay.

The causal meaning of the two subordinators emerged later than the purpose meaning, aligning with the well-attested purpose-to-cause grammaticalization cline (Cristofaro 2003: 161–162; Croft 1991: 293; Heine and Kuteva 2002: 246–247; Zaika 2019: 25; but for a different view on directionality, see Schmidtke-Bode 2009: 177). This development can be seen as semantic bleaching, as purpose is often considered to subsume cause (Boguslavskaja and Levontina 2004: 69–70; Kovačević 1988: 47; Sæbø 1991: 623; Wierzbicka 1998: 184). In this light, the purpose meaning represents the older layer and the causal meaning the newer one. In this perspective, the nineteenth-century loss of the causal use is remarkable in that the earlier meaning survived while the later one disappeared. This development – schematically A > AB > A – contrasts with the typical grammaticalization path A > AB > B, where the intermediate stage serves as a bridging context. My paper focuses on the AB > A stage and explores why the newer causal reading was lost over time.

The paper is organized as follows. Section 2 introduces the relevant subordinators within the broader context of causal constructions in nineteenth-century Russian and provides background on their early history. Section 3 presents mood distinctions and offers quantitative empirical evidence for the decline of the relevant indicative causal constructions; it also discusses semantically mixed uses. In Section 4, I speculate on potential reasons for the discrepancy in the development of purpose versus causal constructions. The main findings are summarized in Section 5.

2 Causal subordinators in nineteenth-century Russian and data collection

The nineteenth century, particularly its first half, saw major changes in the system of Russian causal subordinators (Bulaxovskij 1954: 395–401; Say 2021; Vinogradov and Švedova 1964: 180–188). By the century’s end, the system of causal (and also purpose) subordinators had reached a form that remains largely stable today. For an overview of causal subordinators in Modern Russian, see Belyaev (2015) and Pekelis (2017, 2022).

Changes in the early nineteenth century were shaped by two opposing trends. Several subordinators derived from nouns (blago ‘good’, raz ‘time’) or multi-word expressions (potomu čto ‘along.it that’, tak kak ‘so as’) rapidly increased in frequency and advanced along grammaticalization clines, replacing some formerly common causal markers. Most causal subordinators that fell out of use were Church Slavonic in origin (ibo, poneže, poeliku, zane); their multi-morphemic structure grew opaque and contemporary writers who opposed archaic syntax ostracized them (Vinogradov 1941: 547).

In this context, two complex causal subordinators – dlja togo, čto and zatem, čto – followed an unusual development. Both have a transparent three-part structure: a preposition (dlja ‘for’ and za ‘behind’), a demonstrative in the appropriate case (togo – genitive, tem – instrumental), and a generalized complementizer (čto or čtoby, with by as a subjunctive marker, discussed below). These subordinators fit a typologically common pattern; in Europe, at least, “subordinators of this type are overwhelmingly employed for the expression of non-temporal relations, especially for Cause (e.g. French parce que)” (Kortmann 1997: 221; see also Blümel and Pitsch 2019 on Slavic and German; Zaika 2019: 12 for a broader typological perspective). In Russian, they emerged rapidly around the seventeenth century (Georgieva 1968: 144–145; Vinogradov and Švedova 1964: 115–118) and now dominate the system of causal subordinators, beginning with the default potomu čto ‘because’.

The development of dlja togo, čto(by) and zatem, čto(by) in the nineteenth century was unusual, marked by complex oscillations between cause and purpose meanings and the eventual loss of their causal use. My analysis focuses specifically on these nineteenth-century changes, but to understand them better, some background on the subordinators’ earlier history is needed.

Both subordinators are relatively recent and underwent significant, well-documented changes from their emergence in the Middle Russian period^[2] until the early nineteenth century (see Borkovskij 1979: 229–230, 306–310; Borkovskij and Kuznecov 2006: 496–502).

The purpose uses of both complex subordinators are typically based on čtoby, a marker that was well attested in purpose clauses from the Old Russian period and that eventually emerged as the main dedicated purpose marker. However, throughout the early history of Russian, the mood distinction of čtoby as opposed to čto in the semantic domain of purpose was not perfect (Borkovskij 1979: 224; Borkovskij and Kuznecov 2006: 501). The elements dlja togo and zatem were sometimes used as “extensions” of čtoby, probably from the fourteenth century on (Borkovskij and Kuznecov 2006: 501), but these combinations were relatively infrequent in the Middle Russian period (Borkovskij 1979: 229–230). In Modern Russian, complex subordinators based on čtoby – including dlja togo, čtoby – are commonly used as subordinators of purpose, while the subordinator čtoby itself continues to be the default marker in this domain.

The causal uses of dlja togo, čto and zatem, čto emerged somewhat later than their purpose counterparts – arguably in the fifteenth or sixteenth century – and became frequent in the seventeenth century (Borkovskij and Kuznecov 2006: 497). Unlike the more or less dedicated čtoby, the form čto was always multifunctional. On its own, čto was only rarely used as a causal marker, even in earlier periods, and the use of “extensions” eventually became almost obligatory for expressing causality (Borkovskij 1979: 306–309).

In both cases, constructions with “extensions” are based on anaphoric uses of the demonstrative, with the clause expressing the cause or purpose serving as the antecedent. Such anaphoric uses of togo in dlja togo and even tem in zatem are common in eighteenth-century texts, as well as in earlier ones.

As part of the three-part causal marker, the demonstrative is used cataphorically to refer to the cause expressed in the dependent clause introduced by the complementizer (see Borkovskij 1979: 308 on Russian; and Schmidtke-Bode 2009: 84 on interpreting the main clause element as a correlative placeholder linking the construction to relativization and information structure):

When the three components of the marker appear adjacent, as in (1), the demonstrative’s cataphoric force weakens, paving the way for a unified causal subordinator to emerge (Borkovskij 1979: 309).

The subordinators dlja togo, čto(by) and zatem, čto(by) were fairly frequent around the turn of the nineteenth century. Preliminary counts rank them third and tenth, respectively, among causal subordinators in early nineteenth-century fiction. The latter was much more common in poetry, as in (5), ranking third (Bulaxovskij 1954: 399; Say 2021).

However, both dlja togo, čto(by) and zatem, čto(by) largely lost their causal use over the nineteenth century (Vinogradov 1941: 574), unlike other subordinators of the same type.

Importantly, the two subordinators shared at least one more feature: both showed cause–purpose syncretism around the turn of the nineteenth century. This pattern is typologically widespread (Kortmann 1997: 198–199; Martowicz 2011: 138; Schmidtke-Bode 2009: 152–154). In what follows, I trace how the two conjunctions lost their causal function but retained their purposive function.

I have used data from the Russian National Corpus (https://www.ruscorpora.ru); all numbered examples are drawn from it. I searched nineteenth-century fiction and verse for occurrences of dlja togo or zatem followed by čto or čtoby within the same sentence (queries run in January 2022), then manually removed noise (e.g., adverbial zatem meaning ‘afterwards’). The cleaned samples contained 3,184 and 861 examples, respectively. I annotated all examples for: (i) the exact form of the complementizer; (ii) the presence of punctuation or lexical material between dlja togo or zatem and čto or čtoby. Some subsamples were also annotated for (iii) the semantic type of the subordinate clause (see Section 4).

3 Mood, frequency, and meaning

Both subordinators in my focus can take either čto or čtoby (or its shortened form čtob, a distinction not addressed here) as their third component. These are the two main Russian complementizers, differing in mood: čto is indicative, čtoby is subjunctive. The element b(y) is the regular analytic subjunctive marker, typically combining with the past tense (l-form), such as in On pozvonil ‘He called’ vs. On pozvonil by ‘He would call/would have called’. Thus, čtoby-clauses function as subjunctive complements, with the mood marker fused with the complementizer.

For subordinators with dlja togo and zatem, the čto vs. čtoby contrast strongly correlates with causal vs. purposive meaning – a pattern attested since the emergence of these constructions (Borkovskij and Kuznecov 2006: 496–502). In most cases, čto signals cause, while čtoby indicates purpose; see (1) vs. (2) for dlja togo and (6) vs. (7) for zatem.

Besides the l-forms, čtoby can also combine with infinitives in same-subject configurations:^[3]

The link between purpose and subjunctive clauses, including čtoby + infinitive, is well motivated: by definition, purposes describe events that “must be unrealised at the time of the main event” (Thompson et al. 2007: 250), which is associated with the core meaning of the Russian subjunctive. Systems like nineteenth-century Russian – where cause and purpose clauses share markers but differ in subordinate mood – are common typologically (Thompson et al. 2007: 251), including languages such as English (two types of for-clauses; Van Gelderen 2022: 177–185) and Romance (Sæbø 1991: 623–624).

While the čto vs. čtoby contrast cannot be fully equated with the causal vs. purpose distinction in nineteenth-century Russian (see below), it serves as a clear proxy (Borkovskij 1979: 309); see also Kortmann (1997: 90) on “divergent syntactic constraints” as key to subordinators’ polyfunctionality. Table 1 shows absolute and relative frequencies of čto and čtoby with dlja togo by period; Table 2 presents the same for zatem.

Tables 1 and 2 show that the proportion of indicative subordinate clauses declined in both subordinators during the nineteenth century. However, the decline with zatem lagged behind dlja togo in timing (cf. Vinogradov and Švedova 1964: 184–187), meaning the process affecting dlja togo began earlier and had lower ratios throughout the century. Given the mood–function link, Tables 1 and 2 indirectly confirm the decline of causal uses of dlja togo and zatem throughout the nineteenth century, leading to their near disappearance in present-day Russian.

My next goal is to provide an attempt to support the same claim in a more direct way. One of the major complications here is the fact that, as I mentioned above, the correspondence between the mood and the subordinator’s function was not quite strict. In particular, čtoby was sometimes used in a causal subordinate clause (Klangová 2017: 85; Petrenko 2009: 58) or at least in what looks like a causal clause to a modern reader. In particular, this was typical of negated contexts:^[4]

Dobrushina (2012: 127–129) analyzes uses like (9) as instances of the “epistemic meaning” of by, contrasting them with the much more frequent purpose-related uses in (2), (7), and (8). In any case, uses like (9), where čtoby refers to a situation in the relative present or past (rather than the relative future), became increasingly rare and were almost non-existent by the end of the century. In other words, subjunctive subordinate clauses – which gradually came to dominate – by the late nineteenth century almost invariably signalled purpose.

Deviations of the opposite type, where čto conveys a purpose-like meaning, are also attested, as seen in (10), though they remained rare throughout the century.

Cases like (9) and (10) highlight the difficulty of distinguishing causal from purposive subordinate clauses. It has been known since Aristotle that cause and purpose are semantically close (Thompson et al. 2007: 250–251), with purposes often seen as a subtype of causes (Sæbø 1991: 623). In any language using the same marker for both, the key question is whether this reflects polysemy or vagueness. If the distinction is seen as gradual (Tuggy 1993), the issue becomes where the usage lies on the continuum.

From this perspective, it is not surprising that the two functions – cause and purpose – were not fully differentiated in constructions with dlja togo, čto(by) and zatem, čto(by) in early nineteenth-century texts. As discussed above (especially with examples (9) and (10)), the formal distinction between indicative and subjunctive or infinitive clauses was not yet systematic (see also Klangová 2017: 85; Pekelis 2017: 124; Švedova 1980: 580). This situation was inherited from earlier periods (Borkovskij 1979: 308–309). One can argue that over time, the two usage types eventually crystallized. The key argument in favour of this interpretation comes from uses that, in a sense, occupy a halfway position between cause and purpose – a situation deserving focused discussion.

It is well known that “in the typical case, time reference of the causal proposition is prior to or simultaneous with that of the main proposition” (Sæbø 1991: 625), which distinguishes it from purpose clauses. However, the dependent clause may sometimes assert a real situation – typical of causes – while still containing future-oriented elements. One such case arises when čto combines with the morphological future tense, as in (11).

However, other possibilities exist. A relatively common one is a finite dependent clause headed by a desiderative verb whose subject is coreferential with that of the main clause:

Thus, clauses introduced by čto in sentences like (11) and (12) meet the definitional criteria of causal clauses but, on a deeper textual level, reveal the agent’s intentions – typical of purpose clauses. ‘X is doing P1 because X wants to P2’ can be paraphrased as ‘X is doing P1 in order to P2’.

Based on these observations, I can now present calculations on the types of meanings found in nineteenth-century constructions with dlja togo, čto and dlja togo, čtoby. I have used the same sample and time periods as in Table 1. All instances of dlja togo, čto – a dwindling minority – were manually annotated. I also annotated four random subsamples of dlja togo, čtoby entries (100 examples per period). I distinguished three categories: “cause”, “purpose”, and “mixed”. The first two are straightforward. The “mixed” category includes (i) causal clauses with future-oriented elements (see (11)), (ii) subordinate causal clauses coordinated with clear purpose clauses, and (iii) rare cases where the context made disambiguation impossible. The results are shown in Table 3.

The main conclusion from Table 3 is that the contrast between the distributions of dlja togo, čto and dlja togo, čtoby is already sharp at the start of the nineteenth century and becomes even clearer by its end. However, Table 3 shows only relative frequencies. When combined with the raw data on the overall frequency of these constructions (see Table 1), the findings support the key empirical observation: causal uses of dlja togo, čto(by) declined throughout the nineteenth century. The formal differentiation of moods seems to lag behind the semantic specialization of dlja togo, čto(by) as a purpose subordinator, so even present-day Russian clauses introduced by dlja togo, čto (as in example (12)) can be acceptable if they denote purposes at the textual, if not morphosyntactic, level.

4 Possible explanations

5 Summary and conclusions

In early nineteenth-century Russian there was a plethora of subordinators displaying great variation and overlap in form and function; over the century, the system gained uniformity and functional specialization (Vinogradov and Švedova 1964: 9). My findings fit well within this framework.

The two subordinator families analysed above – based on dlja togo (‘for it’) and zatem (‘behind it’) with either the neutral complementizer čto or the subjunctive čtoby – were partly vague between causal and purpose meanings in early nineteenth-century Russian. Over the century, the indicative form declined sharply, nearly disappearing for dlja togo, čto. Meanwhile, both subordinators became firmly established as purpose markers, combining with an infinitive or, less often, a finite subjunctive clause.

The development described here fits the type where “meanings that at one stage are united (vague) … become separate over time” (Tuggy 1993: 278–279). However, the Russian subordinators’ case is unusual: the older usage (purpose) persisted after a period of syncretism (A > AB > A), whereas typical grammaticalization follows syncretism as a bridge from old to new usage (A > AB > B).

It is difficult to pinpoint specific causes for the loss of the two subordinators’ causal indicative uses and the specialization of their subjunctive forms as purpose markers. However, I hypothesize these changes were driven by analogy with related constructions. First, semantic shifts in the subordinators mirrored those of the prepositions dlja and za, which ceased to be productive causal markers in the nineteenth century. Second, the default interrogative purpose adverbial začem (‘what for’) likely reinforced purpose constructions with zatem, čtoby and may have slowed the decline of zatem, čto. Finally, dlja togo, čtoby and zatem, čtoby filled the niche of segmented counterparts to the simplex purpose subordinator čtoby, a niche unavailable to complex causal subordinators.