TYPOLOGY OF THE REALITY STATUS CATEGORY IN SELECTED LANGUAGES. IS THE HABITUAL IN POLISH AN INSTANCE OF (IR)REALIS OR MODALITY?

The present article is aimed at examining the category of the reality status by discussing the dichotomy “realis / irrealis” in the context of the categories of modality, habituality and futurity. Prototype analysis is juxtaposed with scope analysis, and the category of the habitual is discussed from the typological perspective as well as from the perspective of its connection with the category of futurity. The paper presents aspect diversity of habituals (perfective and imperfective aspect and its contextual implications) as well as the differentiation between the habitual and modality. A special focus is on the prototype analysis and its application instances in Polish, English and Hebrew. The primary objective of the paper is to show that, although it is possible to treat irrealis as notional category, the habituals in Polish and many other Slavic languages (e.g. Czech) should be identified with the modality domain rather than irrealis category. The paper is also an attempt to provide an insight into the distinction between (ir)realis and encoding systems of modalities as the habitual aspect displays modal category features in many languages (including Polish).


Introductory remarks
The category of reality status, often described as realis / irrealis opposition is problematic and its validity is put into question.The literature on the subject lacks the universal definition of the term, not to mention the fact that its very existence is dubious.Most of the researchers agree that although it is not possible to state a synchronic typological category of reality status, the notion can be useful in the bottom-up method of approaching particular languages and language families, i.e. for comparative and historical purposes, as proposed by Ferdinand de Haan (2012).Realis / irrealis is often identified with modality mood, the approach present mainly in the works of Talmy Givón (2001), who coined the notion "mega-modality" including smaller units ("sub-modes"), which all have an underlying denominator called epistemic uncertainty.This approach has also been challenged by the efforts to redefine irrealis in non-modal terms, i.e. treating irrealis as a "supercategory" including three conceptual domains sharing the trait called non-actualization of a SoA (State of Affairs): counterfactuality, non-exclusion of factuality and non-referentiality (Pietrandrea 2012).Habitual aspect is an interesting case to discuss.Apparently, it should be categorized as belonging to the realis domain (if the action is/was habitual, it seems natural to be/have been actualized in the reality), but it becomes clear that many languages mark the habitual either identically to other irrealis categories or display a separate irrealis morpheme for this aspect (de Haan 2012).In the present paper I aim to prove that Polish and many other Slavic languages (e.g.Czech) mark the habitual as modal rather than irrealis.

Reality status -general assumptions, opposing views
The two major assumptions about the reality status are the claim assigning the term autonomous status as a distinct grammatical category and the opposing view describing its nature as coextensive with modality.The former is exemplified e.g. by Elliott (2000) who ascribes the reality status two values: realis (or neutral) and irrealis.The basic distinction is made in the domain of realization / actualization of the SoA.If the latter is asserted by the proposition as an "actualized and certain fact of reality", this proposition is realis.On the contrary, irrealis is the feature of the proposition describing SoA which "belongs to the realm of the imagined or hypothetical, and as such it constitutes a potential or possible event but it is not an observable fact of reality" (Elliott 2000: 66-67).This view results in a great variety of the reality status realization -in some languages it should be obligatorily marked morpho-syntactically, whereas in others it is optional or required only in particular syntactic contexts.One of the arguments supporting the distinction of reality status from modality domain is that the latter is highly speaker-oriented, whereas reality status is perceived as "the grammaticalized expression of location in either the real or some unreal world, its component contrasting terms being realis and irrealis" (Elliott 2000: 67).I will enhance on the relation between realis / irrealis and modality in the next section.
Researchers who support the grammatical validity of reality status have pointed to the alleged "potential actualization" as the decisive factor constituting irrealis, being at the same time the core and shared meaning of all irrealis SoAs (Verstraete 2005).Consequently, irrealis marking of counterfactual and negative SoAs (lacking the alleged "potential actualization") is treated as the result of pragmatic implicature.Parallel term to "potential actualization" is "epistemic uncertainty" (Givón 2001).Both notions refer to the feature which enables the synchronic use of irrealis markers and possibly stimulates their diachronic development (Givón 2001, Verstraete 2005).Although the distribution of the markers across languages seems chaotic, the typological differences between the irrealis patterns originating from diachronic diversity do not preclude their extension to other irrealis SoAs on the basis of cognitive mechanisms.The starting point to any extensive examination of (ir)realis should be the question: "Is it possible to identify cases in which the conceptual distinction between actualized and unactualized situations plays a role in the processes of extension of particular forms from one conceptual situation to some other unactualized situation?"(Mauri, Sansò 2012: 6).Here, the positive answer implies the existence of diachronic developments of given constructions and the possibility of the notion of "(un) actualized SoA" being valid at the synchronic level as well.On the contrary, the negative response puts into question the autonomous nature of reality status and condemns the category to identification with modality domain.Indeed, there are languages which seem to "neglect" the pragmatic aspect of observing the resemblance within different uses of irrealis constructions, followed by diachronic development of irrealis markers.Parallel process took place in the case of the subjunctives, deriving from old present forms which have lost their original contexts as a result of the emergence of a new (progressive) present (Haspelmath 1998).Here the polysemy is not connected with the observed resemblance between the irrealis situation types, but displays "an indirect result of a different grammaticalization path" (Haspelmath 1998: 34).
Discussing the problematic nature of realis / irrealis studies, Ferdinand de Haan (2012) points to three distinct areas which overlap in the procedure: philosophical (extra-linguistic), linguistic and typological.Although the three domains are interrelated, they are most often in conflict with each other.De Haan admits that it is possible to treat realis / irrealis as "notional categories", i.e. to use them in the philosophical context and to apply this meaning to concrete linguistic excerpt.Obviously, such an approach precludes cross-linguistic examination as it is extremely subjective, and does not provide any coherent formal system.One example of such use is the Austronesian language Tsou (Zeitoun 2005), where voice alternations have been described as displaying realis / irrealis relation: "The whole system <...> is based on a modal dichotomy, where the realis contrasts with the irrealis.In the realis, situations are viewed as having occurred or as actually taking place; in the irrealis, they are regarded as having not (yet) occurred, whether or not they may happen.Both in the realis and in the irrealis, events can be seen as 'immediate' <...> or 'remote' (in time and / or space)" (Zeitoun 2005: 279).

Prototype analysis vs. scope approach
De Haan (2012) discusses two most prominent approaches in the linguistic analysis of (ir)realis -prototype and scope methodology.The first abolishes the validity of binary oppositions within the reality status, lending itself to modern cognitive linguistics.Prototype method recognizes the "core" that is identified with the prototypical instance of a given category, and the "periphery" which embodies the area outside the core.Such a generalization is presented in the table, with the core for the irrealis being the upper-left quadrant and the core for realis -the lower-right quadrant.
Respectus Philologicus, 2017, 32 (37) Scope methodology is based upon the assumption that certain languages have distinct semantic scopes for realis / irrealis categories.Then the hierarchy of scope relations is created to justify the cross-linguistic varieties, using either a formal framework or a functional approach.The former is Cinque's Cartography of Features and the latter -Functional Grammar and Role-and-Reference Grammar (RRG).Both approaches assign reality status a position in the hierarchy alongside with the syntactic features.

Realis / irrealis vs. modals
The discussion about the nature of the connection between the two systems should be started from the terminological distinction between the terms modality and mood.The former is used in the semantic domain, whereas the latter refers to grammatical categories (Cristofaro 2012).According to Mats Exter (2012), realis / irrealis domain largely overlaps the domain of modality as well as other domains (e.g.future or habituals).The supporters of the autonomous nature of the reality status category claim that in a top-down approach it is possible to prove that irrealis SoAs are non-actualized, i.e. "presented as not grounded in perceivable reality" (Pietrandrea 2012: 186).
One of the most significant opponents of the autonomous functional role of reality status as an independent grammatical category is Bybee et al. (1994), who claim that the conceptual domain of unreality and the functional domain of modals overlap.Consequently, alleged markers coding irrealis constructions encompass only a sub-set of unreal situations and when one examines the semantic relations between different uses of those markers, it becomes evident that they coincide with the meaning relations within the domain of modality.This idiosyncratic distribution of the alleged irrealis markers is justified by the differences in diachronic patterns affecting the present usage of the morphemes (such as local changes), which transfer grams originally devoted to express agent-oriented modal meanings to those encoding speaker-oriented and epistemic meanings.The latter are generally based on the well-known mechanisms of generalization, metonymic or metaphorical change.Bybee et al. (1994) claim that irrealis should be only considered a "descriptively useful label", as the factor of unreality is not sufficient to justify the shared coding of the sub-functions within the irrealis domain.This approach seems parallel to

Typological approach
As mentioned before, logically habitual aspect should be classified as belonging to the realis domain, as it seems natural that propositions describing actions done habitually should have their reference in real world.However, as de Haan shows in the example of West Grenlandic language (Fortescue 1984, Cristofaro 2004), the situation is often reverse.
The morpheme -ssa is used for (past) habitual aspect but also for obligation and future.The data presents the variety of possible contexts in which irrealis is used, the most common being future and obligation.Habitual aspect here seems to be connected to questions.

Interrelation of habitual and future
In English there is also a visible interrelation of future form will of the verb and habituality in the present:"One can never have enough socks," said Dumbledore. "Another Christmas has come and gone and I didn't get a pair.People will insist on giving me books."(J.K. Row ling, 'Harry Potter and the Philosopher's Stone', chap.12) (Cristofaro 2012).
In Kayardild (Evans 1995: 258-60) the same inflection is used to express both future and habituality, surprisingly in the past.According to Cristofaro, the interrelation of future and habituals may be explained by the fact that as the future is the unknown area, any statements concerning the actions which are going to take place are predictions.The latter are cognitively constructed on the basis of background knowledge, which may encompass the routine -habitual actions.One may infer that "X will do Y" if X habitually does Y.Another explanation is provided by Bybee, Perkins, and Pagliuca (1994: 157) who claim that habitual markers are derived from intentionality markers because what is one's intention or inclination to do will eventually become one's habit.Both hypotheses are of pragmatic nature as they are based upon contextual inferences, which are independent from the factor of unactualization.

Aspect diversity
In Polish, there are two aspectual manifestations of habituality: imperfective and perfective habituals exemplified in ( 1) and ( 2) respectively (Klimek-Jankowska 2012, examples -my own): (1) a. Janek zjada zupę pomidorową na obiad.Janek eat-imp-pres-3sg tomato soup for lunch."Janek eats tomato soup for lunch."b.Moja teściowa gotuje najlepsze pierogi.My mother-in-law prepare-imp-pres-3sg the best dumplings."My mother-in-law prepares the best dumplings." (2) a. Janek pożyczy pieniądze.Janek lend-prfv-pres-3sg money."Janek will lend the money."b.Małgorzata nie wypije taniego wina.Małgorzata not drink-prfv-pres-3sg cheap wine."Małgorzata won't drink cheap wine."c.Franek nie podda się w trudnej sytuacji.Franek not give up-prfv-pres-3sg refl in difficult situation."Franek won't give up in a difficult situation." The difference between perfective and imperfective habituals concerns the "type of evidence on which the generalisations they express are based" (Klimek-Jankowska 2012: 4).Each group of habituals are claimed to be morphological manifestations of two separate kinds of modality.The first type (exemplified in 1) codes descriptive situation-oriented modality, whereas the second type (exemplified in 2) expresses a dispositional (speaker-oriented) habituality.The latter can be expressed by both perfective and imperfective aspects, whereas the former is limited to imperfective aspect only: (1) a. Bożena nie zaakceptuje poglądów swoich przeciwników politycznych.Bożena not accept-prfv-pres-3sg views own opponents political."Bożena will not accept the views of her political opponents."Generally, the dispositional habitual is conditioned by the fact that the truth of its generalization does not necessarily require "enough" instances, which are observable in the real world.The name itself derives from the fact that in the case of dispositional habituals the causal force is a disposition of the subject.What both descriptive and dispositional contexts have in common is that they both share features typical for the habitual, mainly they express "atemporal, law-like, non-accidental generalizations as they are not limited to the actual eventualities, but rather allow us to predict further unexamined cases" (Klimek-Jankowska 2012: 9).

Polish habituals as modals
As Klimek-Jankowska (2012) observes, there is enough evidence to state that Polish habitual belongs to the modality category.Firstly, there are numerous similarities between dispositional habituals and epistemic modals.Most significantly, the accessibility relation in reference to the moment of speaking is in both cases relative (Hacquard 2006 cited in Klimek-Jankowska 2012), e.g.Jan musi być śpiący po nocnej zmianie "Jan must be sleepy after the night shift".Secondly, both epistemic modals and perfective habituals are speaker-oriented (describing speaker's subjective attitude toward the truth of the proposition).In the combination of epistemic modal and perfective habitual statements, the subjective conviction stays oriented towards the attitude holder in the main clause, e.g.Jan uważa, że Małgorzata nie wypije taniego wina Jan believes that Małgorzata not drink-prfv-pres-3sg "Jan believes that Małgorzata won't drink cheap wine".Finally, while epistemic modals appear only in the scope of certain attitude verbs (Anand and Hacquard 2009 cited in Klimek-Jankowska 2012), such as believe, argue, assume and not e.g.hope, wish, command, perfective habituals are also valid in the scope of the same attitude verbs: a. John {believes, argues, assumed} that the Earth might be flat.b. *John {hopes, wishes, commanded} that the Earth might be flat.epistemic modal b.Marcin uważa, że Jan pomoże w potrzebie.Martin thinks that John help-prfv-pres in need."Martin thinks that John will help in need."perfective habitual c.Marcin ma nadzieję, że Jan pomoże w potrzebie.Martin hopes that John help-prfv-pres in need."Martin hopes that John will help in need".Here, the subordinate clause refers to the concrete situation located in the particular moment in the future, so it can no longer be defined as habitual.

Polish habituals as modals in comparison to Hebrew and English
In Polish, English, Hebrew (and Romance languages, which are beyond the scope of this paper), there exists the so-called retrospective habitual, morphologically marked by past-tense auxiliaries: zvykl in Polish, used to and would in English, haya in Hebrew (Boneh, Doron 2009: 1).Retrospective habituals are instances of modal category, and at the same time they describe a temporal period (or interval) finished and as such departed from the time of speaking.According to Boneh and Doron (2009), the fact that although retrospective habituals express actualized habits, they still belong to modality domain is the evidence for the conclusion that "habituality is primarily a modal category, which can only indirectly be characterized in aspectual terms" (Boneh, Doron 2009: 1).
All the languages mentioned above display more than one formal pattern to express the habitual, which can be coded by a simple form (1a-b) or a periphrastic form accompanied by an auxiliary (2a-b, examples from Boneh, Doron 2009: 5): Both Hebrew and Polish use auxiliary verbs in the periphrastic forms.Interestingly, the periphrastic form in Polish may be even more elaborated and take the form of: Here one cannot agree with Boneh and Doron, who claim that "unlike the simple form, the periphrastic one gives a retrospective view on the denoted state, with an effect of a habitual which is felt to be cut off from speech time" (Boneh, Doron 2009: 6).Such an approach would exclude such elaborated, but still periphrastic form as shown in 3.b.Nevertheless, the rule is apparently valid for Hebrew, as exemplified by the utterance: 4. at zoxeret še-'ax-i haya mefarmet li et ha-maxšev?you remember-SF that-brother-my BE.PAST-3SM format-SM to-me ACC the-computer?ma ani omeret?ma pit'om "haya" -hu adayin!what I say-SF why "haya" he still "Do you remember that my brother used to format my computer?What am I saying?Not used to -he still does!" The example above clearly shows that the speaker conceives it improper to use the periphrastic form to refer to the habit that is still actual.

Conclusions
Taking into consideration various theoretical approaches and applying them to Polish it seems more convincing to assign habituals in Polish modal rather than irrealis category features.Obviously, there still exists a possibility to use Ferdinand de Haan's proposal of the irrealis as notional category, the concept of philosophical and not linguistic or typological nature, and to apply such a concept to Slavic habituals.However, I believe there is enough theoretical and pragmatic evidence to treat Polish habituals as prevailingly sharing features with the modality domain.

Table 1 . Marking of irrealis according to Plungian (2005: 138)
. Pietras.TYPOLOGY OF THE REALITY STATUS CATEGORY IN SELECTED LANGUAGES.IS THE HABITUAL IN POLISH AN INSTANCE OF (IR)REALIS OR MODALITY?first category of de Haan's (2012) division (philosophical use of realis / irrealis treated only as notional categories).
-Ssa is not the only marker of habitual aspect, there is also a habitual aspect -ssar(Fortescue  1984: 279-81).De Haan provides a number of examples from various languages, in which irrealis is coded by a habitual morpheme(Cristofaro 2004: 263): P. Pietras.TYPOLOGY OF THE REALITY STATUS CATEGORY IN SELECTED LANGUAGES.IS THE HABITUAL IN POLISH AN INSTANCE OF (IR)REALIS OR MODALITY?