A corpus-based study of aspect: still and already + verb phrase constructions into Spanish

This paper explores till and already + verb phrase constructions, their semantics, and the resources Spanish uses to convey their meanings . The aim is to explore some of the ways in which the grammars of the two languages, En glish and Spanish, encode aspectual transitions. The contrastive procedure is ba ed on the parallel corpus PACTRES, and combines qualitative and quantitative co rpus analysis: the English→Spanish equivalences are qualitatively identified a n their relevance is assessed quantitatively. The similarities and differences ar e analyzed and used as a basis to formulate some empirically-grounded (although neces sarily partial) considerations on cross-linguistic lexis-grammar relations. The resul ts also raise questions about contrast and grammaticalization across the languages.


Introduction
Both English and Spanish convey aspect by means of a variety of different resources, including adverbs. Still and already + verb phrase constructions constitute an interesting English-Spanish contrastive case study not least because of the complexity of the meanings they convey.
Both English aspectuals and Spanish todavía [still] and ya [already] have received ample attention in terms of their lexical idiosyncrasies, for English: van der Auwera 1993; Michaelis 1993and 1996, Krifka 2002Ippolito 2007; for Spanish : Garrido 1991 andFernández Lagunilla and De Miguel Aparicio 1999;Delbecque 2006. When considered crosslinguistically, however, we are, in a way, in a murky area. Previous work (Krifka 2002) has helped reveal that an exclusively lexical approach to these contexts does not necessarily help, because of different solutions in 1 Research for this article, undertaken as part of the ACTRES program, has been partly funded by the Regional Government of Castile and León [LE227U13]. The acronym stands for Análisis contrastivo y traducción English-Spanish "Contrastive analysis and translation English-Spanish". For updated information on results and findings see http://actres.unileon.es/ [Accessed 27 December 2013]. each of the languages or, simply, because of the absence of formal markers in one of them.
This paper sets out to explore still/already + verb phrase contexts in English and the resources Spanish uses to convey their meanings. The aim is to find out whether Spanish offers additional grammatical alternatives to todavía ['still'] and ya ['already'] + verb phrase constructions which are capable of conveying the aspectual transitions.
An outline of the aspectual semantics of still and already + constructions and their Spanish counterparts is provided in section two below and there is a concise account of the corpus in section three and of the procedure followed in section four. Both the quantitative and qualitative descriptions of these aspectual constructions are dealt with in sections five and six, while section seven is devoted to the contrast proper. The conclusion in section eight, which focuses on the relationship(s) between cross-linguistic contrast and single-language grammatical treatment, points at regularities in grammaticalization and/or "optimization patterns" in Spanish.

Still and already + verb phrase constructions and their Spanish counterparts
The (lexical) semantics of still and already indicate that they are primarily aspectual operators that, associated to the (grammatical) meaning of verbal structures, contribute to refining the expression of the aspectual transitions in a given proposition, in other words, they improve our understanding of the 'internal temporal constituency of the situation' (Comrie 1976: 4). Basically they convey two core notions: 'absence of change or continuation' in the case of the still contexts (example a) and 'imminent change or inception' in the case of those marked by already (Garrido 1991), as in example b. a. They still love you, Neely (FGJ2E.s274) b. My mask was already getting misty (FPT1E.s1121) 2 They also mark, at varying degrees, the attitude of the speaker towards the situation or event (ter Meulen 2004:245) and mark a transition-in the case of already−or a continuation-in the case of still−in the temporal development of the situation.
Still + verb phrase constructions indicate that the situation continues to obtain at the time referred to, and involve some contrast with what might have been expected in a similar situation. These constructions carry a presupposition that the situation obtained before the time referred to−the previous phase−and continues to obtain at the time of reference. They may also involve a contrast with what would have been expected.
Already + verb phrase constructions highlight the time of inception, of the beginning of the situation, indicating that it obtains at the time referred to, and marking a polarity transition from a negative to a positive phase. Again, they may also imply that the situation has begun earlier than expected (Huddleston and Pullum 2002:710-13).
In their aspectual uses, Spanish todavía/aún and ya are taken to be the formal equivalents of still and already -and also of yet and anymore. 3 Todavía + verb phrase constructions are taken to be inclusive, and indicate that the action/event already obtained before the time referred to. In this sense they convey continuity with a previously existing situation, which is generally interpreted as indicating durative aspect, as in example c.
By contrast, ya + verb phrase constructions are exclusive, meaning that the situation obtains at the time referred to but not before, which is generally interpreted as indicating inceptive aspect. It is the case of Victor's expectations in example d. These Spanish constructions also 'combine factual and subjective information about what is happening' (ter Meulen 2004:241), which determines to a large extent which resources are better equipped to convey the information. As in English, the presuppositions concerning the transitions (ter Meulen 2004:246;García Fernández 1999: 3153) from a negative to a positive phase or the continuation of a previously existing one seem to be the key to how we interpret these contexts.
In both Spanish and English constructions aspectual functions and positive/negative polarity marking can overlap. 5 This second function, however, goes beyond the scope of this paper. Rather, this study will concentrate on two issues: -Are these aspectual transitions conveyed by the same lexical−grammatical means or in Spanish are there alternative, equally and/or more frequently used resources? -Do still/todavía + verb phrase constructions and already/ya + verb phrase constructions show comparable degrees of grammaticalization cross-linguistically? Does 'more frequent' imply 'more grammatical'? To try to (tentatively) answer these questions, and some more that arose during the process, a contrastive procedure has been employed using parallel corpus P-ACTRES 2.0. as a source of empirical data.
3. The P−ACTRES 2.0 corpus P-ACTRES 2.0 (http://actres.unileon.es/inicio.php?elementoID=12) is a do-it-yourself (DIY) bidirectional parallel corpus comprising original texts in one language, English or Spanish, and their translations into the other. It contains over 4 million words in both directions. P-ACTRES EN→ES contains slightly over 2.5 million words distributed into different subcorpora: books, fiction and non-fiction, press -newspaper and magazine articles-, and miscellaneous texts. Excerpts of around 15,000 words have been used for the book subcorpora, and full texts for 5 Spanish does not have differentiated resources for positive and negative contexts. Ya and todavía are considered as polarity items by the standard grammar of Spanish (RAE 2009(RAE : 3636-3637 and 2330(RAE -2333 and appear in positive environments. Negative environments use ya no and todavía no respectively. Empirical data show both operators enter a double polar opposition: todavía/ ya no[still/ anymore] vs ya/ todavía no [already/yet] (Sánchez López 1999: 2598-2603 all the others. The version of P-ACTRES ES→EN used for this study dates from March 2013, and contains just under two and a half million words, distributed among books (fiction and non-fiction) press (newspapers and magazines) and miscellanea (table 1). The tagged parallel texts have been aligned at sentence level and can be searched with the Corpus Work Bench browser (CWB) (Izquierdo, Hofland and Reigem 2008). P-ACTRES 2.0 is used to obtain basic diagnostic information about language use and cross-linguistic interpretation. 'Diagnostic' here means 'instrumental' in that translation performance is not the goal of the study but a means to empirically unveil problem areas between English and Spanish and to focus on formally dissimilar equivalents which might be overlooked otherwise. This is also the reason why original Spanish materials (P-ACTRES ES→EN) have not been sampled and analyzed in this particular study.

Procedure
This research combines qualitative and quantitative corpus analysis and is modelled on the following stages: selection, description, and contrast. The P−ACTRES EN→ES parallel corpus querying capabilities which supported our selection criteria include detailed POS selection of up to three items on complete, start, end or part of words (e. g. verb + base form; past tense; gerund or present participle, etc.) among other features. 6 These made it possible to concentrate on those constructions where aspectual transition marking is likely to show, excluding other uses of already and, particularly, of still. 5. 'Still' + verb phrase construction corpus data 5.1. Selection and sampling An initial query for any construction which included still came to 1280, of which 613 (47,9, nearly 50% of the total) combined with verb phrase constructions. This selection restriction was adopted so as to focus on the aspectual uses of these constructions. As mentioned above, P−ACTRES EN→ES search tools offer the possibility of querying for different verbal combinations so as to cover all possible scenarios. To ensure the representativeness of the selection quantitatively, the strategy employed has been stratified random sampling. Table 2 shows that over 25% of the occurrences happen in the environment of an -ing form, followed at a considerable distance by past (15.7%) and past participle (PP) forms (10.7%), and more neutral tenses such as the simple present (9.5%). Constructions including base forms of the verbs ([would, may, etc] still + V base) range between 9% and 6.1%. Every other combination ranks below 5%. Despite the relatively low figures for some of the queries, every single possibility will be considered for cross-linguistic solutions as some of these options may throw some light on the interpretation of these constructions.
The search for translation solutions yielded a marked preference for the periphrastic solutions (table 3 in bold). The periphrastic cluster Seguir/ continuar/ pasar + gerund [continue + ing form] constitutes (40%) of the cases. Additionally, (todavía) estar + gerund and negative periphrases sin dejar de/ no terminar de + infinitive [(not) stop + ing form] contribute a further 7% of the Spanish options. An alternative to this overwhelming preference is the imperfect tense accompanied by aún/todavía [still] (21%) followed by the same construction in the present tense, which represents 20% of the solutions.
All other solutions are below 5%, which may not be important from the quantitative point of view, but do seem quite promising qualitatively.  Gómez Torrego 1999: 3323-3347). A further function of Spanish periphrases is to allow the speaker or writer to present his/her point of view concerning the process, action or event referred to. The case of (mainly) seguir + gerund [continue + gerund] is commonly associated to aspectual content, to the expression of durative and iterative meanings. 8 To interpret seguir + gerund constructions access to the previous phases of the process or state referred to is required. By itself, seguir + gerund indicates that whatever the process, action or state, it was already happening in a previous stretch of time and persists at the time of reference in the text.
As shown in examples 1 to 3, the periphrastic construction can be conjugated in all tenses, according to needs (Matte Bon 1995: 162), and is not accompanied by any additional element that would replicate English still. Just two instances have been found where seguir + gerund is reinforced by todavía (examples 4 and 5). In both cases todavía acquires an additional pragmatic meaning concerned with speaker's (or writer's) expectations: that the situation persists longer that it is reasonable to expect. Seguir + gerund generally conveys an uninterrupted process, either continuous or iterative, and occasionally may alternate as a translation option with no/sin dejar de + infinitive [(without/not) stop + gerund].
Dejar de + infinitive is another grammaticalized periphrastic construction used to convey two possible meanings (RAE: 2178-2180): a) interruption of a state or event (which is not necessarily completed) and b) continuation, when accompanied by a morphological negative, as in example 6: According to Matte Bon (1995: 163), who acknowledges the close relationship of these constructions with todavía /ya (no) constructions, this is particularly true of no/sin dejar de + infinitive, although there is a difference in focus: whereas the periphrastic construction indicates absence of change or interruption of some process or state by means of lexical semantics, todavía relies on the pragmatics of the situation. In this sense, no/sin dejar de + infinitive would be interpreted semantically whereas todavía would have to refer to the contextual information obtaining from the situation. Estar + gerund is another of the periphrastic choices to translate still + verb constructions into Spanish. Unlike progressive constructions in English, estar + gerund does not have tense status in the grammar of Spanish (RAE 2009(RAE : 2185(RAE -2192. 9 It is very frequent in contemporary Spanish and it is extremely sensitive to the grammatical context. Basically, it signals progressive aspect and focuses specifically on the actualization of the internal time of the action, process or event, as in example 7, where the process is presented as ongoing and open-ended. However, estar + gerund may also indicate iteration or gradual change within the process, as in example 8, which conveys the intermittent manifestations of activity and the constant modification of the situation (Yllera 1999: 3402-3412;RAE 2009RAE : 2189

Imperfect + 'aún/todavía'
The second most typical solution is an imperfect tense accompanied by todavía/aún. The Spanish imperfect signals a process or situation continuing in the past with no end-point. All our occurrences include aún or todavía in the construction, indicating that the situation continues to obtain at the time referred to, contrary to what might have been expected. No difference has been reported in the use of the periphrasis or the imperfect construction on the basis of verb semantics (action, event, state…). Neither have corpus data yielded any noticeable systematic difference in the use of todavía and aún, as illustrated in examples 9 and 10,

Present + 'aún/todavía'
Spanish uses the simple present to mark actualization of the process, action, state or event within the framework regarded as here and now by the speaker (Rojo and Veiga 1999:2900-2918, Rabadán 2009. If the context requires presupposed information about the stage prior to the one that obtains, Spanish will make use of todavía or aún to indicate that the situation referred to is a continuation of some state, process or event previously in existence. Examples 11 and 12 illustrate this use but their pragmatics are different. In 11 the situation is neutral in that 'I still have not run out of coffee', whereas 12 must be interpreted that whatever is being referred to was previously made with sugar everywhere, but now this situation obtains only in those places where sugar cane continues to be affordable. Together with these three preferred translation solutions, our corpus search has yielded a number of less frequent choices which, although not relevant quantitatively, may shed some light on the internal workings of both languages, and/or on susceptible points of language interference in this particular area.

'Aún/todavía' + conditional/ future tenses
Conditional and future tenses represent 4% of the translation solutions and in all cases are instances of "third code language" (Øverås 1998), 10 as in example 13, where a present subjunctive [sigan protegidas/ estén protegidas] would be more acceptable.

Present + Ø
Simple Present without any further modification accounts for 3% of the solutions into Spanish. In these cases there is a meaning change in the texts, as in example 14−the English text refers to the continuation of a situation that was widespread in the past, but has become unusual in the present; the Spanish translation just states a present day fact, without any link to a previous situation.

Others
In those cases where the still construction has been reinterpreted lexically into Spanish, as example 19, some degree of meaning loss or change applies. Our study reveals that Spanish favors periphrastic resources over adverbial solutions to convey the aspectual values encoded by still constructions in English, while corpus data show that other options are possible though significantly less used. These findings are discussed in more detail in section seven below.

'Ya' + verb phrase construction corpus data 6.1. Selection and sampling P−ACTRES EN→ES yielded 332 cases of contexts including ya
[already] combined with verbal constructions. 11 Table 4 shows that 45% of the occurrences happen in the environment of a Past Participle form, followed at great distance by -ing verbal contexts (19%), passive constructions (9%) and past tenses (9%). All other possible contexts represent under 3%. Our search of P−ACTRES EN→ES shows that ya is present in over 90% of the translated texts, irrespective of the translation solution (table  5). It also shows a clear preference for the present and past perfect in Spanish−21% and 19% respectively−independently of the input construction in English. As alternatives, our empirical data yield ya + present tense (14%), ya + imperfect (13%), verbal periphrases (12%) and ya + attributive construction (10%). All other solutions are less than 5%, which, while not impressive quantitatively, could be interesting qualitatively.

Description
By contrast with the empirical results for the still constructions, already environments tend to find a much more defined match in Spanish positive phase indicator ya. Ya contexts convey the idea of change into/ inception of a phase or stage that did not obtain before, but that has just started to obtain.

'Ya' + present perfect
The Spanish Present Perfect indicates result in the present of an action started in the past. Our data show that as an equivalent for already constructions, this tense is overwhelmingly accompanied by ya, as illustrated in examples 20 and 21. The pragmatics of the context can be described as 'this process/event is a change with respect to what has come before and applies to the speaker's present'.
(22) This experiment is in progress as I write, but the participants have already told me a great deal about this agony and the stages of despair the rejected lover must endure. (EHF1E.s489) (23) En el momento en que escribo estas líneas, el experimento está en pleno desarrollo, pero los participantes ya me han contado muchas cosas sobre esta angustia y las fases de la desesperación por las que debe pasar el amante rechazado. (EHF1S.s477) (24) … maybe, just maybe, the elderly human might be sacrificed because he had already lived a full life, would not be missed, and had little future. (EGJBM1E.s149) (25) … quizá, sólo quizá, la persona mayor debería sacrificarse porque ya ha vivido su vida, no la van a echar de menos y no tiene mucho futuro por delante. (EGJBM1S.s124)

'Ya' + past perfect
All our Past Perfect cases (19%) include ya in the construction, independently of whether it is a conditional or a subjunctive−sensitive context (example 22), or a neutral factual statement, as in example 23. As its English counterpart, the Spanish Past Perfect encodes an action that has happened before another past time action. Up to an additional 5% of solutions correspond to highly grammaticalized periphrastic constructions, which occupy an important slot in the Spanish system as they assume fine meaning distinctions, often aspectual and/or modal. Our empirical data reveal that ponerse/ empezar/ comenzar a + infinitive [begin to + infinitive], together with estar + gerund are options into Spanish for already+ verb phrase constructions.
Ponerse a + infinitive [start to + infinitive] conveys the sudden inception of an action (Gómez Torrego 1999: 3374), with a certain degree of speaker disengagement as in example 31, where no end or solution to the argument with the waiter is inferred (44) …he was already arguing with the waiter, who pointed at me.
(46) Walking beside her, Langdon was already feeling the tingle of anticipation that accompanied his face-to-face reunions with great works of art. (FBD1E.s989) (47) A su lado, Langdon ya empezaba a notar ese cosquilleo de impaciencia que siempre le invadía momentos antes de ponerse frente a las grandes obras de arte. (FBD1S.s925)

(Ya) + simple past
Spanish Simple Past constructions are perfective by default. When used with ya, as in example 33, they mean "there and then, at that particular point", and ya functions as a deictic, as mentioned in note 2 above. In this temporal use the difference with example 34, which does not include ya, is imperceptible.
(48) Out of the corner of my eye I already knew the challenge came two men down, but I stood in front of three long enough to agree with my earlier assessment. (FSA2E.s753) (49) Con el rabillo del ojo ya vi que el desafío estaba a dos hombres de distancia, pero me detuve delante del tercero el tiempo suficiente para confirmar mi anterior afirmación. (FSA2S.s722) (50) Poppy zippered his coat, eyeing me, and seemed already hunched against the cold outside. (FHC1E.s193) (51) Poppy se subió la cremallera de la cazadora y pareció encorvarse contra el frío de la noche-. (FHC1S.s189b)

Modal constructions
A very small number of solutions opt for constructions combining modal adverbs and non-factual tenses, i.e. some verbal form in the subjunctive or a future, as in example 35. The stance is different in the English and the Spanish texts: the former assumes that the receiver is in possession of that knowledge; the latter conveys a high degree of uncertainty about whether the receiver has the necessary information. 12 Pragmatically, the Spanish text has a certain whiff of superciliousness about it, as it triggers the interpretation "I do not believe that you know".

Others
This scant undefined group consists of those cases where the grammatical already + verb construction has been translated mainly by lexical solutions. Example 36 uses a fixed prepositional phrase but fails to target the inception meaning contributed by already, example 37 has reinterpreted the aspectual transition into a noun phrase indicating 'outcome', and example 38 uses a verb and also misses the aspectual part.
(54) My mind was already racing, developing a new list of what had to be done and done now. (ECR1E.s97) (55) Mi mente iba a mil por hora, reelaborando la lista de los pasos que había que seguir en ese momento. (ECR1S.s101) [My mind was buzz-ing….] 12 The construction posiblemente sabrá is not widely used in Spanish; rather it seems to derive from "third code" (translated) uses (Øverås 1998 Corpus data reveal that the aspectual values conveyed by already constructions are mostly encoded in Spanish by means of ya + tense combinations. Aspectual periphrases are a not a mainstream option and frequently include lexical aspect and/or are reinforced by adverbs. A more detailed discussion follows in section seven below.

Contrast
Corpus data show markedly different scenarios in the resolution into Spanish of the English still/ already + verb phrase constructions. Still solutions present two major strategies, verbal periphrases and aún/ todavía combined with a tense form. The (slight) preference for periphrastic constructions (47%) suggests that these forms constitute a widely accepted resource to encode the aspectual core meaning of 'continuation of previous stage or actions, which persists at the time of reference'. The one and only function of Spanish verbal periphrases such as seguir/ continuar + gerund is precisely to express a number of finely grained aspectual values that are frequently encoded by lexical means in other languages, such as French (e.g. Il chante toujours/ Sigue cantando [He is still singing]) (Yllera 1999:3394). This specialization would explain why the aspectual operators aún/ todavía are not required in these constructions, as corroborated by our findings-over 90% of the seguir/ continuar + gerund cases do not feature either of them.
By contrast, the majority of estar + gerund constructions incorporate aún/ todavía to clearly delimit the aspectual meaning of 'continuation of stage or action at the time of reference', as in examples 7 and 8 above.
Dejar/ terminar de + infinitive stands in a complementary relation to seguir/ continuar + gerund, i.e., stating one implies negating the other. Our corpus data show that both periphrases are used in conjunction with negative elements sin [without] and no in order to maintain the aspectual meaning of continuation.
The second strategy (45%) amounts to reinforcing tenses that have the feature of 'continuation, duration' inbuilt in their grammatical meaning with aspectual operators aún/ todavía. These include primarily the Spanish imperfect and the present tense.
Other options (8%) include lexical solutions, particularly lexical aspect, and the use of litotes. Particularly interesting is the latter, statistically a marginal use illustrated by example 18 above, but that supplies an alternative option: the Spanish translation keeps the aspectual transition -the situation continues to obtain at the time of reference-but the means of expression are, to a point, converse [still packed →unpacked yet].
Corpus data show that already + verb phrase constructions have been translated overwhelmingly by means of formally parallel solutions into Spanish, that is, tenses accompanied by ya (71%). The choices have been the present and past perfects (43%), the present (14%), the imperfect (13%) and the simple past (45) as a minority solution.
Verbal periphrases are not a prime option here, as they were for the still + verb phrase constructions. Considered jointly, (ya) empezar /comenzar/ponerse a + infinitive and (ya) estar + gerund, account for just 12% of the solutions.
Empezar/comenzar a + infinitive is considered a 'phasal periphrasis' (RAE 2009(RAE :2172(RAE -2175 and may indicate 'initial stage of an activity or state within a situation' or contribute an iterative or habitual slant when the initial stage applies to an indeterminate number of occurrences of the same activity, e. g. Empecé a ver películas antiguas [I started watching old films]. Ponerse a + infinitive, roughly equivalent to the previous periphrasis, focuses specifically on the inception point, without signaling any progression. An additional distinction is that inception is a built-in feature of empezar and comenzar as aspectual verbs (García-Miguel 2005: 405-411). Corpus data reveal that ya is frequently included in the periphrastic construction as a way to signal that the inception is underway at the time of reference.
As discussed above, estar + gerund concentrates on the progression of an action or situation, and is frequently accompanied by adverbs that focus on a particular point of the progression. Ya signals that the situation being referred to did not happen earlier.
Attributive constructions translating English non-finite structures account for a further 10%. Solutions featuring the Spanish (perfective) Simple Past convey temporal, not aspectual meanings, and represent 4% of the cases analyzed. The rest of the options consist of modal solutions (4%) that shift the meaning partially in Spanish and lexical combinations (3%) that also miss the aspectual phases highlighted in English. P−ACTRES EN→ES data reveal an overwhelming majority of solutions focus on the translation of aspectual transitions into Spanish. Corpus data suggest that the complementary relationship that holds between English constructions containing still and anymore seems to be replicated in the Spanish periphrases seguir/ continuar + gerund and dejar/ terminar de + infinitive. No such relation has been observed between any two of the solutions given to already + verb phrase constructions.
Whether these intra-linguistic relationships may contribute to more clearly defined cross-linguistic equivalences needs to be further explored on a broader basis.

Concluding remarks
Our experimental study shows the existence of formally dissimilar equivalents in Spanish for the aspectual English constructions. It also raises a number of issues, some contrastive, some related to grammaticalization.
The aspectual transitions encoded by still + verb phrase constructions can be encoded in Spanish by means of verbal periphrases indicating precisely how the actions, processes and events are seen in terms of their relation to the previous stage and the expectations of the speaker concerning the subsequent stage. Spanish aspectual periphrases follow regular patterns, since they can be conjugated in all tenses (appropriate secondary interpretations ensuing), are strongly lexicalized and show high degrees of grammaticalization 13 : they form a grammatical, phraseological and semantic unit, they cannot be interpreted in terms of their individual components and are associated to certain contexts by speakers of the language. All of which qualifies them as cross-linguistic equivalents of the English construction.
According to corpus data still + verb phrase construction finds significant translation equivalents in seguir + gerund and (NEG) dejar de + infinitive, as alternatives to 'aún /todavía' + verb phrase constructions. In over 90% of these periphrastic solutions the adverb is not present, which suggests a considerable semantic strength. In the few cases where it appears, it adds a hint of annoyance on the part of the speaker, who would rather have the situation changed. Our findings suggest then that verb phrase constructions with todavía belong in well-defined aspectual transition (phasal) uses which may involve some degree of emphasis.
Already + verb phrase constructions primarily prefer a more straightforward solution, ya + verb phrase constructions, to indicate the aspectual transition: that the situation obtains at the time of reference and is assumed to continue. Empirical findings suggest that ya + verb phrase constructions mark the change of period (RAE 2009(RAE :2331 whereas (ya) + verbal periphrases contexts focus exclusively on the inception (and/ or iteration) point.
Parallel corpus data highlight the importance of tackling jointly the study of both positive and negative resources English-Spanish (Bosque 1980, González Rodríguez 2009. Hence, scalar analyses of English aspectuals, among them Israel (1997 and2011), do not serve well in a joint study involving ya and todavía, as the set of relations represented by already and yet on the one hand, and still and anymore on the other, are not found in Spanish (or at least not in the same way).
These results underscore more general contrastive issues, such as the fact that the grammatical treatment of any given phenomenon in the different languages may introduce an additional contrastive parameter in the analysis. English features specialized separate forms for positive and negative contexts, Spanish does not. The analytical framework(s) that work(s) for the English aspectuals involve(s) polarity. Where English offers 'polarity-sensitive aspectual adjuncts' (Huddleston and Pullum 13 Spanish grammar makes a distinction between fully grammaticalized verbal periphrases (RAE 2009(RAE : 2140(RAE -2222, as it happens here, and semiperiphrastic constructions (RAE 2009(RAE : 2125(RAE -2131 2002:710), Spanish gives 'phasal adverbs' (RAE 2009(RAE : 2330, with polarity regarded as unrelated to the main topic and-in a way-marginal. Any attempt at bringing both angles together in the analysis requires the analyst to carry out a "grammatical treatment contrast" prior to the case study. Another contrastive issue concerns the benefits of contrasting (grammatical) constructions rather than single elements, as it helps reveal cross-linguistic associations that are not readily accessible.
Contrastive data also draw attention to whether the higher frequency of verbal periphrases over aún/todavía + verb phrase constructions points towards a distinctive grammaticalization pattern in Spanish (Mair 2004) that may convey the aspectual phases involved in the interpretation of these contexts more accurately and 'grammatically'. If we are to take "semantic suitability, salience and frequency" (Traugott and Heine 1991:9) as indicators of grammaticalization processes, Spanish aspectual periphrases do qualify as an expressive optimization over other resources: seguir, continuar, pasar + gerund are aspectually similar to constructions with verbs mantener(se) and quedar(se) [keep doing]. These do not, however, qualify as periphrases. 14 Another, equally interesting question arising from this study is why the periphrastic solution is substantially less favored as an equivalent to already + verb phrase constructions.
To answer these questions further work is required using Spanish materials translated into English as well as non-translated materials. The former will help identify additional cross−linguistic equivalences and, possibly, spot lexical-grammatical choices that can be put down to translated language (Johansson 2009:12). Non-translated materials will serve as control data to verify whether or not grammatical differences between original and translated resources are statistically significant. Pertinent results can then be fed into the PETRA 1.0 © application for use in the evaluation of the grammatical quality of translated texts (Rabadán et al. 2013).