Particle

A particle is an ad-rhematic commentary. This succinct definition implies at least four properties of particles. First of all, particles are naturally analysed in the context of the themerheme structure, rather than the clause structure described by the syntax of dependency, in

which particles do not participate. In consequence, on the one hand, particles should be examined in a produced sentence (utterance), and thus taking into account prosodic and contextual factors. On the other hand, however, their non-involvement in the clause structure means that no particle is required by any component of the clause and the particle itself does not require any specific element in terms of grammar or meaning. Nevertheless, the commentary it contributes must be consistent with what is being said. As a result, a particle that would weaken a speaker’s conviction, such as chyba, has no place in a sentence that conveys a strong conviction of that specific speaker (e.g. Jestem o tym przekonany.) There are no reasons, however, for it not to appear in a sentence that informs us of the convictions of someone other than the speaker himself (On chyba jest o tym przekonany.), since there is no conflict between the two (non-identical) epistemic states of a single person. Secondly, participating in what is being said, a particle is a marker of the attitude of the speaker with regard to the knowledge being conveyed. Thirdly, the commentary conveyed by the particle concerns a specific element of that knowledge – what is being said about something, and thus the rheme. These last two properties enable us to identify the basic functions of particles. They include: 1. informing the recipient about the degree of certainty of the speaker with regard to that what precisely may be said about a given object; 2. reporting the choice of saying one thing rather than the other about a given object; 3. a reflection connecting what is being said about a given object to saying something else (primarily saying the same of a different object or saying something else of the same object). And fourthly, since a particle remains connected to an element of the utterance, i.e. the rheme, it is natural for it to precede the rheme (the postposition, or – in other words – moving the particle to the end of an utterance, that is the position which is naturally stressed in a sentence, is not out of question for particles, on condition that the stress is linearly moved to preceding elements. Nevertheless, in a vast majority of cases this position would be a marked one). Given that the rheme is set out with the use of prosodic measures, a particle may be linearly moved away from it, e.g. Chyba zjadłem już pięć ciastek, może nawet sześć. Some particles (for instance epistemic ones) are more mobile from this perspective, while the mobility of others (e.g. exemplifying particles) is much more constrained.

1. Properties of particles

From the outset, research on particles in their contemporary understanding has focused on the two properties of units belonging to this class: their function and their asyntactic quality. We disregard here the traditionally identified group of morphemes (-że, –ć, –ś, –li, –kolwiek, –no, –to, –co, –li) and units currently classified into various different classes (bądź, by, oby, czy, niech, nuże, bodaj) that textbook grammars refer to as particles (in Polish: wyrazki)

(cf. Szober 1923, 1957, 1966; Klemensiewicz 1937, 1965; Lehr-Spławiński, Kubiński 1946; Gołąb, Heinz, Polański 1968; Jodłowski 1971). Early on, particles became recognized as a tool used by the speaker to steer the speech process, i.e. to express speaker’s attitude towards the content of the clause being uttered and to indicate to the interlocutor “from what perspective the sentence is to be interpreted” (Mirowicz 1948, 1949). This role of particles has found its reflection in the alternative names given to this class: additional intellectual marker (Klemensiewicz 1953; Misz 1968), modalizer (modal expression, modalizator in Polish) (Laskowski 1984, 1998) or modulator (modulant in Polish) (Jodłowski 1976), cf. also clause modifiers such as oczywiście, głównie or tylko involved in broadly construed modality (Grzegorczykowa 1975).

From the perspective of syntax, particles are described as invariable (uninflected)^* non-accommodated units (Klemensiewicz 1953; Saloni 1974; Grochowski 1986, 1997; Wiśniewski 1995; Wróbel 1996, 2001; Laskowski 1998), which are not syntactically independent (Saloni 1974; Laskowski 1984) and remain in a specific syntactic relationship with regard to the remaining elements of a sentence. They have been noted to have the capacity to bind either to an entire clause or to “any of its constituents, forming a portion thereof” (Klemensiewicz 1965; Misz 1981). Scholars have pointed out their variable, unlimited distribution (Misz 1968, 1981; Makarski 1971; Grochowski 1983; Laskowski 1984; Wróbel 1996), sometimes emphasizing their capacity to enter into relations with a specific class of parts of speech (especially nouns, cf. Laskowski 1998; Grochowski 1986, 1997; Wróbel 2001), which puts them in opposition to other classes of parts of speech (adverbs). Scholars have also emphasized their freedom in terms of order (Mirowicz 1948; Laskowski 1984; Grochowski 2006, 2016), which by some (Grochowski 1983, 1984, 1986, 1997, 2006) is associated with the abovementioned unlimited distribution, with reservation that a change in the linear position triggers a change in the reference of the particle. However, parallel research considers particles as elements external to a clause, independent from its structure (Klemensiewicz 1937, 1953; Mirowicz 1948;

Jodłowski 1976) and not classified as its constituents, belonging to the “framing of a sentence” (Misz 1981). In consequence, the description of particles from the syntax perspective has become highly complex (cf. comments on particles-adverbs in Saloni 1974).

* Importantly, in the oldest – extremely vast – understanding of particles, this group was identified with an uninflected word (cf. Małecki 1863; Malinowski 1869; Ulitzka 2009).

The thesis that particles do not perform a determining function and are not constituents of the grammatical structure of a sentence has been clarified in the latest approach to metatextual units from the perspective of the theme-rheme structure (Misz 1981; Wajszczuk 1997, 2005; Grochowski, Kisiel, Żabowska 2014; Grochowski 2015), where they are seen as synsyntagmatics units. This approach clearly places the classes of units at this level beyond the syntax of dependency. Since particles are not involved in a clause structure, they cannot fill the rhematic (nor the thematic) part, which means that they are non-contrastively unstressed and cannot be negated (Grochowski 1986, 2015; Kisiel 2012). This property explains the comments that particles have no impact on the correctness (Misz 1981) or the “communicative core” of a sentence (Jodłowski 1976). Secondly, as placed beyond the clause syntax, they do not fill any connoted positions and they do not require any element from a syntactically and/or semantically specific class (Wajszczuk 1997; Wróbel 2001; Grochowski 2003, 2006). Nevertheless, since classes and subclasses of metatextual expressions are internally hierarchical (Żabowska 2009), some particles may fill positions opened by other particles (Grochowski 2015), cf. Ma chyba tylko brata. This approach allows us to clarify the manner in which particles participate in the speech process as comments on rheme. Initially, researchers (e.g. Misz 1968, 1981; Grochowski 1983, 1984, 1986, 2006) classified particles, together

with order and prosodics (though secondary with respect to the latter, cf. Grochowski 1984) as elements actively creating the T-R structure, i.e. defining the rheme. The double role of particles as both the determinant of and the comments on the rheme posed a risk of creating a vicious circle in the description of the relations between the units from this class and the T-R structure (Kisiel 2009; Żabowska 2014). In consequence, particles were eventually removed from the set of the exponents of rheme. Meanwhile, their positional flexibility may be explained by the prosodic marking of the rhematic part (the connection between particles and the stressed element being the “bearer of the content dominant” that was already identified by e.g. Misz 1981).

The concept of particles as comments on a rheme exacerbated the opposition between units belonging to this class and – commonly confronted with them – → adverbs. Since adverbs connote and are connoted by verbs, they belong to the sentence structure and operate at a different organisational level than particles. At the same time, the clear distinction between the two classes triggered research on metapredicative units (Bałabaniak 2007; Danielewiczowa 2012), and thus on the further differentiation of the various meta levels. On the other hand, however, the description of particles from the perspective of the T-R structure prompted necessary comparisons between this class and other metatextual classes, including → ad-rhematic conjunctions, ad-rhematic thematizers and connectors, as well as metatextual comments referring to the entire T-R structure. Insofar as the earlier syntactic descriptions distinguished between particles and conjunctions on the basis of, respectively, having the connective function or the lack thereof (Saloni 1974; Grochowski 1984, 1986, 1997; Maldijeva 1995; Wiśniewski 1995; Wróbel 1996, 2001; this view was criticized by Wajszczuk 1997), the description of both classes from the perspective of T-R structure allowed for identifying further, fundamental differences between units from both classes: while particles bind with one element, mostly the right-sided one (Grochowski 2006, 2015, 2016), conjunctions bind with two; on top of this, the position of particles is variable (Misz 1981), while conjunctions have a fixed position at the beginning of the [constituent] clause (Wajszczuk 1997). At the same time, despite clear criteria, the distinction between particles and conjunctions remains troublesome, as evidenced by the constant shifting of the entire groups of units between the two classes (see e.g. ‘improper’ conjunctions such as skoro, ponieważ, chociaż in Wajszczuk 1997, classified as a subgroup of linking particles in Wajszczuk 2010).

In Polish linguistics, there is very little commentary on the lack of the independent meaning of particles, most likely motivated by the absence of syntactic independence (Milewski 1965; Szober 1957). The prevailing view is that the specific function performed by the units from this class calls for a different approach to their definition than the one applied to the units from the representational level. Anna Wierzbicka (1969, 1971) was among the first to point out this fact, emphasizing the role of metatextual units in constructing the “second line of text” and their function in revealing the speaker. In her work on the semantics of human interaction (1991), she argues that particles, as – quoting John Locke – “abbreviations for whole sentences” and representatives of the action of the speaker’s mind, should be defined through the reconstruction of the sentence they contain in themselves. The thesis on the link between the metatext and the T-R structure developed by Jadwiga Wajszczuk clarified any doubts as to the fact that the definition of particles must refer to the speaker and speech (Grochowski, Kisiel, Żabowska 2009, 2014).

2. The scope of the class

Since particles were identified early on as the operators of the spoken text with a modal-focusing function, expressing the attitude of the speaker to the content and its sentimental,

volitional and logical modulation (Mirowicz 1948; Szober 1957; Pisarkowa 1978), it comes as no surprise that one of the most frequently studied particle subgroups are epistemic particles (Żabowska 2008). Their clear distinctiveness is confirmed by the fact that, even when described as adverbs, they were still classified as a single group (formal adverbs in Krasnowolski 1906; mood adverbs Krasnowolski 1909; Gaert ner 1933, 1938; modal adverbs in Grzegorczykowa 1975). Importantly, most of the units exemplifying the class of particles in older studies – except for the very few, such as tylko, nawet or właśnie – would today be classified precisely as epistemic particles (Mirowicz 1949; Klemensiewicz 1965; Laskowski 1984). It was not until a systematic study of particles undertaken by Maciej Grochowski that it became clear that “the expression of modality is not […] a specific property of particles nor is it their only property” (Grochowski 1984: 78).

We are unaware of any attempt at any subdivision of particles where epistemic particles would not hold a leading spot. Henryk Gaert ner (1938) distinguished adverbs of judgement (niechybnie, naturalnie, podobno, może) which he classified as the adverbs of mode, along with the adverbs of emphasis (zwłaszcza, nawet), inclusion (także, w dodatku) and restriction (jedynie, przynajmniej, tylko). When it comes to additional components, Zenon Klemensiewicz (1953) distinguished components of assessment (oczywiście, zapewne, chyba), emphasis (tylko, nawet, aż, już) and reference (po pierwsze, z jednej strony, następnie, ponadto, przede wszytkim). Stanisław Jodłowski (1949) contrasted valorizing particles (in Polish: waloryzanty) modulating the logical value of utterance (owszem, chyba, bodaj, bynajmniej, but also nie, oby, czy) with the situating particles (in Polish: sytuanty), placing the content of an utterance in a context without specifying or detailing it (tylko, jeszcze, nawet, zwłaszcza). In his subsequent classification (Jodłowski 1971), the scholar distinguished modal modulants (na pewno,

może, podobno, chyba, oby, niech), which mark the kind of speaker’s attitude towards the reality of the content of the utterance, valorising modulants (owszem, naprawdę, istotnie), which describe the relation of the content of the utterance to reality, situating modulants (tylko, jeszcze, właśnie, przeciwnie, zwłaszcza), which place what is being said in a broader context, affective modulants (nawet, zresztą, przecież, przynajmniej, na szczęście, raczej, aż, no) that contribute the emotional layer to the utterance and the modulants introducing justification or reinterpretation (na przykład, ewentualnie, mianowicie). When it comes to modalizers, Roman Laskowski (1984) distinguishes adverbs (może, niby, chyba, istotnie) expressing speaker’s attitude to the uttered sentence, pragmatic operators changing the set of presuppositions implied by the sentence (jeszcze, właśnie, tylko, zwłaszcza), and quantifying modalizers (prawie, ledwo, całkiem). In his semantically motivated classification, Maciej Grochowski (1986) distinguished the following subclasses of particles: 1. expressing the attitude of the sender to the logical value of the communicated judgment, namely 1a. his or her knowledge (istotnie, oczywiście, wprost, poniekąd, wcale nie) or 1b. sender›s or someone else›s beliefs and suppositions (na pewno, prawdopodobnie, podobno) with regard to the subject, and 2. particles that do not speak directly about the attitude of the sender towards the logical value of the communicated judgment, but communicate 2a. the frequency of occurrence of a certain state of affairs (na ogół, zwykle), 2b. time sequence (dopiero, już, wreszcie), 2c. the existence (też,

jeszcze) or non-existence (tylko, jedynie) of other objects with properties possessed by the object referred to in the sentence, 2d. being one of the many objects that the sender could talk about (na przykład, choćby), 2e. the dominance of a given object (przede wszystkim, zwłaszcza). By contrast, the subdivision of particles in the SGPP (Grochowski, Kisiel, Żabowska

2014) relies on the distinction between the tiers of actual speech affected by the commentary conveyed by the particle, i.e. 1 knowledge as to the truthfulness of what we say (epistemic particles); 2. choice of what we say (particles commenting on speech); 3. embedment of what

we say in what has been said (conclusive particles); 4. comparison of the objects of which we speak (comparison particles); 5. linking new information to the information already given (complementing particles). Each of the groups (nodes) thus distinguished has been further subdivided on the basis of the semantic criteria.

In Polish linguistics, neither the understanding nor the scope of the class of particles has been fixed. Currently, units such as niech, bodaj or oby, which used to be referred to as particles in traditional syntax, are classified as mode operators requiring a specific grammatical form of the verb (Grochowski 1997; Wróbel 2001), while czyż, a nuż, no are included in the class of declarativeness modifiers. Other units which impose categorical restrictions on the component with which they co-occur, such as ad-numerative operators (niespełna, około, ponad cf. Duszkin 2010; Doboszyńska-Markiewicz 2011, 2013) and ad-substantive operators (wprost, byle, lada) should also be excluded from the class of particles. By contrast, appositions (in Polish: dopowiedzenie) – a class whose distinctness has been challenged from the

very beginning (Grochowski 1984) – should be regarded as a specific type of particle use (Laskowski 1984; Grochowski 1986, 1987; Dobaczewski 1998).

Finally, the inclusion of linking particles (Wajszczuk 1997), also referred to as contextual particles (Walusiak 1999) or ad-clausal operators (Grochowski 2000, 2003), such as zaś, jednak, przecież, wszak, into the class under discussion remains questionable. In traditional syntax and older structural descriptions, they are viewed as conjunctions (Klemensiewicz 1937; Jodłowski 1976; Saloni, Świdziński 1998) or a subclass (e.g. of adverbs in Krasnowolski 1897) with a linking function. In studies that rely on the linking function-based distinction between the classes of particles and conjunctions, such lexemes are included into one of them,

yet the decisions of specific scholars with regard to a particular unit may vary (cf. e.g. jednak as a conjunction and przecież as a particle in M. Grochowski (1997) as opposed to both units being classified as particles by Bańko in 2001 and in the SGPP). The key difference between linking particles and the “classic” ones lies in their occurrence in complex utterances. Should we assume that they operate on entire clauses / T-R structures/ more than one element of two T-R structures (Grochowski 2000; Kisiel 2012, 2018) or that they open two syntactic positions (Walusiak 1997; Grochowski 2000), ascribing any of these properties to them would eliminate them from the class of particles (cf. also Grochowski 2009; Ulitzka 2009; Wajszczuk 2000; Grochowski, Kisiel, Żabowska 2014; Żabowska 2014).