English non-finite participial clauses as seen through their Czech counterparts

Based on the data from a parallel English-Czech corpu s, the present study offers an analysis of 600 English Ving participial clauses through their Czech translation correspondences, divisible into less and more expli cit types. The less explicit Czech counterparts highlight the analytic character of En glish either in cases where the translation counterpart is synthetic (i.e. merging the meaning of the finite verb and the participle into one verb) or where the participle r esembles, in its function, a preposition. The more explicit (i.e. finite-clause) Czech counter parts attest to the backgrounded information status and semantic indeterminacy of th e English participial clause. Instead of an expected tendency to render their meaning in Czech by a similar, syntactically subordinated, structure, namely dependent clauses, it i the simple coordination that appears to represent best the semantic indeterminac y of the relation of the English participial clause to its superordinate element.


Introduction
Participial forms as a means of condensation are employed to a different extent in English and in Czech. In English the nonfinite V-ing forms appear both in an adverbial and a postmodifying function frequently, and they are not marked stylistically (e.g. I lay on my bed, tossing restlessly and We passed a sign pointing to the village, respectively). In Czech, non-finite transgressive forms can be used for the adverbial function. Transgressives have properties similar to participles (dívka zamávala usmívajíc se na něho -the girl waved at him smiling) but, unlike participles, cannot function as modifiers of nouns. They are archaic and highly formal. The postmodifying function can be expressed by deverbal adjectives (dívka usmívající se na něho -the girl smiling at him).
The comparison of participial forms in English and Czech as two typologically distinct languages (one of which uses participles as condensers systematically, while in the latter a corresponding form has become obsolete, almost non-existent) consequently invites a twofold corpus analysis: first, a systematic overview of the translation correspondences, and second, observations as to whether any potential additional lexico-grammatical and functional aspects of the English participles may be pointed out when contrasted with their Czech counterparts. The first step in the analysis catalogues the means (forms and syntactic structures), while the second draws attention to some more general aspects that are usually not attributed explicitly to English participial clauses.

English and Czech
The long tradition of applying the synchronic contrastive functional approach in Czech English linguistics (since the late 1920s, cf. Dušková 2012: 21-26) 1 highlights the differences between what may be called nominal (or, verbo-nominal) as opposed to verbal ways of expression, in English and Czech, respectively (He gave a squeak. (NP+VP+NP) as opposed to Czech Zapištěl. (VP)) (cf. Mathesius 1975, Renský 1964a. The description of these differences is concerned e.g. with the question of the noun-verb quotient (i.e. the frequency of the respective word-classes) in the two languages (Renský 1965), with changes in the word-class in translation (Poláčková 1988) or with syntactic constancy in translation (Dušková 2012). Such studies show that the verbo-nominal ways of expression can be seen as a feature concomitant with an analytic type of language, i.e. they are typical of English as compared to Czech, which is dominantly a synthetic language with rich inflection. From this perspective the tendencies in the use of nonfinite forms, as items with nominal properties, 2 can be expected to differ in the two languages as well.
Drawing on extensive previous research, Dušková (2012: 27) distinguishes three types of verbo-nominal means of expression: "(i) nominal tendencies in primary denomination […], and two types in the sphere of syntax: (ii) total suppression of the finite verb (as a component of a more general tendency to condensation, also called grammatical nominalization), and (iii) verbo-nominal constructions with semantically empty verbs (dissociation of verbal predication into the categorial and notional components, also called semantic nominalization)".
Contrasting English and Czech, the first type (i) can be exemplified by have influence -působit ('to-influence'), feel the need -potřebovat ('to-need'); the third (iii) by She gave the chair a gentle turn. -Po-otočila křeslo. ('gently-turned-she the chair'). The second (ii) type of verbonominal ways of expression, which is of interest in this study, includes English participial condensers.
Typological differences between English and Czech verbs can schematically be summarized as follows: English finite verb phrase is typically compounded (analytic), aspect is not grammaticalized, predicates display a tendency to semantic nominalizations (to have a chat). These features contribute to what has been described as reduced dynamism of the verb. The reduced dynamism of the English verb is evident also in the use of sentence condensers, i.e. the participial forms, which express only relative temporal meaning and voice, and syntactically operate as non-finite (participial) adverbial and postmodifying clauses providing supplementary background information (Biber et al. (1999), but cf. also Fuhre (2010) for a detailed discussion).
Czech, on the contrary, as a synthetic type of language, is characterized by rich inflection, contributing to the strong dynamism of the verb as the categories are expressed in the lexical finite verb form and include aspect. Non-finite transgressive forms (comparable to English participial adverbial clause but not to a postmodifying clause, e.g. dívka, umývajíc nádobí,… ('the girl, washing up, ...') are highly obsolete; de-verbal adjectives with a similar condensing function express 2 "The difference in the verbal and nominal expression between Czech and English is manifested in English syntax by the fact that English employs indefinite, i.e. nominal, forms to a much larger extent and in its own way." (Mathesius 2001: 64, translation is ours) gender, number and case concord: dívka umývající nádobí, ... ('the girl washing up ...').
As pointed out by Dušková (2012: 25) " [Mathesius] does not deny the English verb its basic function to express predication -English naturally expresses predication primarily by the verb -but as compared with other languages, it often displays verbo-nominal constructions where other languages have verbal forms of expression." Moreover, "as shown by contrastive studies examining other points and aimed at other goals, differences in verbal and verbo-nominal means of expression between English and Czech keep manifesting themselves as a major, non-negligible feature conducive to various types of divergence." (ibid: 27) Participial condensers, displaying a high ratio of divergent translation counterparts (over 90 per cent, cf. Table 2), seem to represent this tendency par excellence.

Material and scope of the study
This study focuses on adverbial and postmodifying V-ing participial clauses, comprising present and perfect participles, both subjectless and with an overt subject (i.e. absolute constructions). Adverbial participial clauses introduced by conjunctions (i.e. augmented participial clauses) were also included in the data (see Table 1). De-participial secondary prepositions and conjunctions, such as according to, assuming that, concerning, including, regarding, were noted but not included in the data set since they can no longer be considered adverbials or modifiers. The data were drawn from the synchronic multilingual parallel corpus InterCorp, which is being compiled at Charles University in Prague (Čermák and Rosen 2012). The corpus comprises 31 languages, with Czech as the pivot language, i.e. it contains bidirectional translations between each of the languages and Czech. Where the same text has been translated between Czech and several languages, multilingual comparisons of the translation counterparts are possible. The texts include fiction, whose alignment is checked manually, and the so-called 'collections' of journalistic and legal texts aligned automatically. The current size of the corpus (March 2013) is 138.8 million tokens in the foreign language texts in the core fiction part, and 728.5 million tokens in the 'collections'. The present study relies on the fiction texts of the English-Czech section of the corpus. 3 The analysis is based on 600 translation pairs of English V-ing participial clauses and their Czech counterparts. A subcorpus of six English novels, three British and three American, and their translations was used (see Sources). From each novel, 100 participial clauses and their Czech counterparts 4 were included in the data set. The overall quantitative results are presented in Table 2.

The counterparts of English participial clauses
The Czech correspondences of English participial clauses were classified primarily into congruent and divergent counterparts (Johansson 2007: 23-26). The congruent correspondences comprise Czech constructions which display the same degree of syntactic compression as the English participial clauses, namely Czech transgressives and de-verbal adjectives. Overall, only 9.5 per cent of the counterparts can be considered congruent. This is mainly due to the low representation of congruent correspondences among the counterparts of the English absolute and adverbial constructions. There is no congruent counterpart of the English absolute construction available in Czech, and the adverbial non-finite constructions (transgressives) are obsolete and rare. English postmodifying participles, on the other hand, are frequently translated by congruent counterparts, i.e. by de-participial adjectives.
Differences in translation preferences between participial adverbials on the one hand and postmodifiers on the other can be observed throughout the whole translation paradigms of the two constructions. English postmodifying participial clauses tend to be translated by modifiers within the noun phrase: postmodifying de-verbal adjectives or relative clauses (42.1 and 40.4 per cent of counterparts, respectively); English adverbial and absolute participial constructions display a preference for coordinate finite clause counterparts (cf. Table 2). The individual types of correspondence will be dealt with below, focussing on what the Czech counterparts can indicate about the functions of the English participial constructions.
The only transgressive forms that exceed the frequency of one per million words in the Czech synchronous corpus SYN2010 6 are those which have become grammatically invariable, having changed their function to that of secondary prepositions or adverbs, such as počínaje ('starting with'), nemluvě ('notwithstanding'), konče ('up to'). The Czech de-participial secondary prepositions can serve as counterparts of certain English participles (počínaje in example 2). The same process, namely the reanalysis of the participle as a preposition ("decategorialization", cf. Hopper and Traugott 2003: 108), operates in English, as illustrated by other Czech secondary prepositions occurring among the counterparts of English -ing forms (concerning corresponds to the preposition ohledně in example 3).
As a result of the restricted set of the archaic transgressive forms, the degree of congruence of the counterparts of adverbial participles is very low (2.1 per cent). 7 On the contrary, the second type of Czech congruent counterparts, i.e. de-verbal adjectives, are used to render 42.1 per cent of postmodifying participles into Czech. Like English participles, the Czech de-verbal adjectives are capable of expressing the contrast of voice and (relative) tense, but they are tied to the head noun more closely than the English postmodifiers through concord in number, case and gender (the adjective vedoucích in example 4).

Divergent correspondences
The congruent translation choices being highly restricted, other types of Czech counterparts have to be sought, especially for the English participles in adverbial function.

Verbless counterparts
A relatively infrequent option (10.7 per cent of counterparts overall with a similar representation among adverbial and postmodifying participles, 10.6 and 12.3 per cent, respectively), consists in employing a verbless construction in Czech. The English participle either disappears in the Czech translation or is translated by non-verbal means, namely prepositional, adjective, adverb or noun phrases. As a result, the Czech sentence displays a higher degree of condensation than the source one, at the cost of becoming less explicit due to the absence of the verb -both lexically and grammatically -since the verbless construction cannot express tense and voice contrasts. One of the verbless translation options consists in using an elliptical construction in Czech (cf. Karlík et al. 2012: 467): in example (5) the unexpressed verb can be recovered as the transgressive maje / majíc / majíce ('having'), expressing 'possession' or 'belonging' in general (cf. the Czech verbo-nominal expression mít oči upřené na něco ('have one's eyes fixed on something') and its English verbal equivalent stare at something) (Malá and Šaldová 2012).
(5) Harry went back to the kitchen, still staring at his letter.
The Czech phrasal counterparts can be illustrated by examples (6) to (8).
The group comprises zero correspondences of the participle (e.g. sounding half exasperated corresponding to napůl podrážděně ('half exasperatedly') in example 6). The English participial adverbial clause is rendered into Czech as an adverbial expressed by an adverb phrase; the Czech adverb mirrors the complement of the participle lexically. As suggested by the synthetic Czech correspondences, the above constructions can be seen as a manifestation of the analytic nature of English. The verbal meaning is dissociated into two components: the finite verb conveys the grammatical functions and the general categorial meanings of direction or duration, while the participle constitutes the semantic core of the construction. Czech, as a predominantly synthetic language, relies on affixes for the expression of grammatical and categorial meanings (cf. also the eventive object constructions give it a try -zkusit to, or go and fetch st. -přinést něco, come to an end -skončit, come true -uskutečnit).
We have also included in the synthetic group of correspondences sentences with direct speech, such as example 13. What they have in common with the other synthetic constructions is that the finite verb of reporting (usually say) has no overt counterpart, and the participle corresponds to a finite verb in Czech. The Czech translations show that the punctuation marks may be sufficient to indicate the boundaries of direct speech. In English, however, a reporting finite clause with the verb say -although semantically redundant -seems to be preferred to bridge the gap between the direct speech and the attendant activity expressed by the participle. The lexical meaning of the participle and the 'reporting' function of say merge in the Czech finite verb counterpart.

Finite subordinate clauses
Although adverbial participial clauses are generally thought of as expressing some type of adverbial modification, and as such can be expected to correspond to subordinate adverbial clauses systematically (Dušková 2006: 583), the data show that in translating an adverbial participle into Czech coordination prevails (example 14), the ratio of hypotaxis and parataxis being 1 : 4. Postmodifiers, on the contrary, are mostly rendered into Czech by subordinate adjectival clauses (7.5 : 1), as in example 15.
Mrs. Dursley came into the living room carrying two cups of tea. Do obývacího pokoje vstoupila paní Dursleyová a přinesla dva šálky čaje. Lit.: Into living room came Mrs. Dursley and she-brought two cups of-tea.
On the one hand, the Czech counterparts in the form of a subordinate clause (example 16) retain the hypotactic relationship between the dependent clause and an element of the superordinate clause. On the other hand, these Czech counterparts are more explicit than the English participial clauses due to the finite form of the predicate (uvědomil si -'he-realized' in example 16), the overt expression of the subject (or at least its easier identification through the categories of person, number and gender, e.g. the suffix -l in uvědomil si indicates, apart from the past tense, third person, singular number and masculine gender), and -in adverbial clauses -the conjunction signalling not only subordination but also the semantic relationship between the two clauses (protože -'because' in example 16). Subordinate finite clauses therefore do not seem to be strictly equivalent to the participial ones since they reduce the semantic indeterminacy typical of the participial clauses (cf. Kortmann 1991).
A mere juxtaposition of an adverbial clause and its superordinate clause, without a subordinator explicitly marking the relationship, is a peripheral means of clausal connection in Czech. Such sentences are formally identical with those whose clauses are in a paratactic relationship but semantically the inter-clausal relationship can be described more adequately as subordination. In our data it is represented by one translation pair only, example 17.
(17) She was enjoying herself immensely, having taken a long shower, eaten a pound of candy, and watched the TV nonstop. Měla se nadmíru dobře, pořádně se vysprchovala, snědla kopu cukroví a vytrvale se dívala na televizi. Lit.: She-was enjoying herself immensely, thoroughly sheshowered, she-ate a lot of candy and she-watched the TV.
Unlike the English sentence where anteriority is indicated by the form of the participle (having taken/eaten/watched), in the Czech translation the causal interpretation rests on the reader's general knowledge of the world, supported merely by the perfective aspect of the verbs vysprchovala se and snědla ('she-showered, she-ate'). This translation solution seems to be facilitated by the coordination relationship within the adverbial element. The translation of a participial clause by a finite adverbial clause always involves making the potential semantic relationship between the participial and its superordinate clause explicit. As pointed out by Kortmann (1991: 116), "the range of interpretations that the addressee views as being available to a given free adjunct/absolute is restricted and identical with the set of interpretations that emerges when all instances of intersubjective variation for this adjunct/absolute are taken together". The translator's task, then, involves selecting merely one of these potential interpretations. The participial clause in example 18 is potentially ambiguous between causal, temporal simultaneity/accompanying circumstance, and possibly even concessive interpretation. By using the conjunction protože ('because') the translation narrows down the options to 'cause' only.

(18)
You got a wealthy lawyer from a wealthy firm deliberately allowing a wrongful eviction to occur, and as a direct result my clients got tossed into the streets where they died trying to stay warm.
The inexplicitness of the semantic relationship between the clauses may be retained in the Czech translation if a non-restrictive relative clause can be employed as a counterpart of the English adverbial clause (example 19). These postmodifying clauses "convey supplementary information about the referent of the noun. Since they are irrelevant for the description of what the speaker has in mind, various semantic relationships may obtain between the content of the relative clause and that of the superordinate clause, such as a causal relationship" (Karlík et al. 1995: 497-498, translation ours).
Generally, adverbial clauses with 'less informative' semantic roles 10 of accompanying circumstance, temporal simultaneity or manner tend to be translated by means other than finite adverbial clauses more frequently than the 'more informative' adverbial clauses (Mašková 2013: 59). The semantic relations expressed by finite adverbial clauses in our data were most frequently causal (i.e. clauses of reason, purpose, or result; example 18), temporal (example 20), and conditional and concessive (example 21). ... they reached the rock, where Uncle Vernon, slipping and sliding, led the way to the broken-down house. ... dorazili k útesu, kde je strýc Vernon, i když se smekal a klouzal, dovedl k polorozpadlému stavení. 10 Kortmann (1991) uses the term 'informativeness' to indicate how much co-/contextually substantiated evidence or general knowledge is needed on the part of the reader/listener to identify the semantic role of a participial adjunct. More informative semantic relations, such as concession, condition or causal relations, require "a considerably higher amount of knowledge or evidence" (Kortmann 1991: 121).
Lit.: ... where them Uncle Vernon, even though he slipped and slid, led to the broken-down house.
For English postmodifying participial clauses, the subordinate finite clause appears to be the preferred translation correspondence, although even here the finite clause is more explicit than the participial one. In the Czech translation in example 22, for instance, "the aspect ... is specified as perfective although the original allows both perfective and imperfective readings" (Mašková 2013: 63), i.e. while in English coming may indicate an action either simultaneous with or anterior to being deeply committed to volunteer counseling, the translation eliminates the simultaneity option.

Finite coordinate clauses
While postmodifying participial clauses tend to be translated into Czech by finite relative clauses, the adverbial ones display a marked preference for finite coordinate clause correspondences. 11 Both types of Czech 11 The prominence of the paratactic correspondences of participial clauses was related to "more deep-reaching" differences between the two languages by Vachek: "[o]bviously, one has to do here with something more deep-reaching than a mere difference in syntactical forms: what is involved is two different ways in which the two languages tackle the realities of the outside world. In Czech one observes the tendency to dissociate the reality to be expressed into a number of actions or processes, which may be mutually either coordinated or subordinated; in English, on the other hand, a different tendency is at work, viz. one that endeavours to grasp the same reality as a single, basic action or process, counterparts involve a higher degree of specificity due to the overt expression of grammatical categories in the finite verb predicate. At the same time, both make it possible to render the underspecified semantic relationship between the participial and the superordinate clause into Czech. The relative clauses, moreover, retain the same syntactic function as the postmodifying participial clauses. The coordinate clause correspondences highlight "the only allimportant difference" between paratactic and hypotactic clauses: "Unless special lexical means (e.g. connective adverbs) are employed, propositions encoded via parataxis are of equal informational rank; in other words, none of the propositions can be marked (!) for presenting background information. It is at least this piece of information, i.e. the presentation of one proposition as backgrounded, which always gets lost in paraphrases of free adjuncts/absolutes by means of coordinate clauses." (Kortmann 1991: 113) 12 Therefore, while the semantic indeterminacy is retained in the Czech coordinate clause counterparts of participial adjuncts, the structuring of information within the sentence becomes more opaque. The non-finite form of the English participial adjunct marks the information it conveys as concomitant or of lesser importance. "The backgrounding, dedynamizing effect of the participial form is lost but the relative degree of [communicative] dynamism of the coordinated clauses may be indicated by other means than in the English sentence, namely by linear modification." (Malá and Šaldová 2012: 153-4) The clause which corresponds to the participial adjunct can be shifted to the sentenceinitial position to indicate its being a part of the thematic layer of the sentence (example 23).

(23)
No doubt I might have taken up with her straightaway the matter of the following day, but she and Ichiro did not stay on the veranda, going inside to wash their hands.
absorbing all other potential actions or processes as its elements or concomitant circumstances." (Vachek 1955: 65) 12 Cf. Mathesius (1966Mathesius ( [1942: 80-81): "Most frequently the task of the transgressive is to express an action concomitant with the main action and temporally so equal to it that both constructions can easily be swopped depending on which action is considered the main one by the writer." Nepochybně bych s ní byl otázku programu na příští den probral na místě, ale zašli si rovnou umýt ruce a na verandě se nezdrželi. Lit.: ... but they-went straight to-wash hands and on theveranda they-did-not-stay. Finite clauses are -due to the presence of the finite verb predicate and the subject -more explicit than the participial ones. The postmodifying finite clauses, however, make it possible to retain two important characteristics of participial clauses, namely their backgrounding effect (both participial and finite clauses are modifiers of the head noun) and the indeterminacy of the inter-clausal semantic relationship. In adverbial finite clauses subordination is retained but the obligatory subordinator resolves the semantic indeterminacy of the relation between the adverbial and the superordinate clause. The fact that 63.4 per cent of adverbial participial clauses are rendered into Czech as coordinate clauses suggests that the possibility of keeping this semantic relation inexplicit seems to be so important that it is often given preference over the syntactic expression of dependency relations in the translation.

Conclusion
The present paper set out to explore the properties of English V-ing participles as they can be seen through their Czech translation counterparts, focussing primarily on the corresponding 'units of meaning'. The analysis of the correspondences, however, has also led to some observations concerning the differences between the two languages in general. The divergent counterparts can be divided into less explicit -"shorter" ones (4.2.1 and 4.2.2) and more explicit -"longer" ones (4.2.3 and 4.2.4). The merger of the superordinate verb with the participle in what we have termed 'synthetic correspondence' attests to the synthetic character of the Czech language as opposed to the analytic tendencies manifested in the structure of the predicate in English. Conveying directional, temporal or aspectual modification, this type of Czech counterparts is associated mostly with adverbial uses of English participial clauses. The verbless correspondences, on the other hand, reveal a class of semantically weak English participles, such as relating to, comprising, using, leading to, looking. They may be used to modify a noun (postmodifiers) or a verb (adverbials) but their syntactico-semantic function is limited to serving as a linking element used to tie the modified element with its elaborator. They may be seen as functionally equivalent to prepositions, as demonstrated by the Czech prepositional counterparts. Taking a step further, the participles may become fully reanalysed as secondary prepositions. Accordingly, the de-participial prepositions can also be translated by Czech prepositions which have come into existence through a parallel process of decategorialization from transgressives.
Apart from these correspondences, quite rare in fiction, the only congruent counterparts, namely de-participial adjectives, are found among the translations of postmodifying participles. However, these counterparts are more explicit than the English participles since the tie between the head noun and the postmodifier is signalled not only by the position but also by concord in number, case and gender between the noun and the postmodifying adjective.
Comparing the two languages, the tendency in English to use a relational (i.e. semantically weak) participle to express syntactic relations in the sentence overtly may be required by the limited freedom of word ordering. This principle can be illustrated by the following example (though from a different set of data) with a discontinuous noun phrase, which is not necessary (and impossible) in Czech, whose word-ordering matches the information structure: (28) … on ordinary days a notice used to hang in the window, saying: Umbrellas Recovered. … ve všední dni visívala v okně tabulka s nápisem: Potahování deštníků. Lit.: ... on ordinary days used-to-hang in the window a notice with the inscription: Umbrellas Recovered.
The choice of a Czech finite-clause counterpart is inevitably connected with explicitation in translation due to the overt expression of the subject and the verbal categories. The two types of finite-clause correspondences -a hypotactic and paratactic one -each point out a different characteristic of the English participial construction. The range of subordinating conjunctions used in the translations attests to the diversity of inter-clausal semantic relationships that participial clauses can convey; the explicitness of the subordinate-clause counterparts, on the other hand, contrasts with the indeterminacy typical of the participial clause. In coordinate-clause correspondences, the semantic inter-clausal relationships do not have to be made explicit. The possible nonequivalence between the participial clause and its finite counterpart concerns the structuring of information. The subordinate non-finite clause makes it possible to present a proposition as 'backgrounded'; in coordinate clauses other means, such as the linear arrangement of the clauses, have to be sought to indicate the information structure.
Although restricted in its scope the present study hopes to have shown that a look at the English V-ing participial constructions through their Czech correspondences may reveal some of the functional, semantic and textual properties of participial clauses. At the same time, the contrastive view highlights the problematic areas of overlap between adverbial and postmodifying participial clauses, the status of absolute constructions, word-class shifts or semantic indeterminacy.