Adverb

1. Adverbs in the general classifications of parts of speech; basic semantic, syntactic, and morphological properties

In the oldest, mostly semantically-oriented typologies of parts of speech, adverbs were usually described as names denoting the properties of actions, states or properties themselves. As a result, they do not relate directly to objects, but rather to properties assigned to objects, which means that in a sentence they are either the determiners of a verb ([idzie] szybko, [przyszedł] dzisiaj), an adjective (wyjątkowo [ciekawy], lekko [ciepły]) or another adverb (całkiem [dobrze], mało [rozsądnie]). In Stanisław Szober’s “Gramatyka” (1957 [1921]), adverbs are defined as signs for the objects of thought, and within this group they indicate (broadly construed) properties. The group further includes pro-adverbs (tam, którędy, kiedy…) as well as numeral adverbs (dwojako, poczwórnie, pięciokrotnie…). Unlike the two last classes, adverbs are signifying words (unlike the deictic [pro-adverbs] and signifying-deictic [numerals]). Just

like the signs for objects (nouns) and the properties of objects (adjectives), but unlike the signs for relations (prepositions, conjunctions, and particles), adverbs are syntactically independent words. Henryk Gaert ner (1938) also classifies them as independent parts of speech, additionally describing twenty semantic subclasses distinguished on the basis of the “type of circumstances” (adverbs of place, time, manner, degree, quantity, multiplicity, frequency, repetitiveness, order, effect, specifying, restricting, etc.). Zenon Klemensiewicz (1937, 1960) also views this class as designating the properties of activities, states or other properties, and,

in addition, specifying place and time. He distinguishes it from the class of “additional markers” such as zwłaszcza, przynajmniej, tylko, aż etc. In the classification by Tadeusz Milewski

(1965), based on the principles of semantic connotation, adverbs are tertiary naming words, alongside primary nouns and secondary verbs and adjectives. In this approach, adverbs describe secondary expressions, connoting them semantically and syntactically. Their deictic equivalents are pro-adverbs, while their ordering equivalents are adverbial numerals. They are also distinct from non-independent syntactic components (particles). Stanisław Jodłowski

(1976), who explicitly declared the use of combined classification criteria, defines adverbs as words with a denoting function that are mnemonic, non-naming, and specify a property of an activity or of a property. From the perspective of syntax, they may actively perform the function of an adverbial, while passively they take adverbials themselves. Jodłowski further distinguished them from the various types of – according to his terminology – modulators (modulant in Polish), as the elements of the “framing of an utterance,” which do not engage in basic structural relations. However, since the number of detailed studies on such extra-structural elements of sentences was initially low, the grammars from that period subsumed them in a group that was not only heterogeneous in itself (serving as – according to one of the researchers – a “trash bin”), but also poorly differentiable from adverbs. The initial attempts at organising the linguistic system in this respect treated adverbs as a sort of supercategory, cf. e.g. the several semantic-distributional classes in the study by the Polish scholar Irena Bellert (1977), who, on the grounds of English linguistics, described adverbial categories according to the logical dependencies that they express. The subsequent interest expressed by researchers in the non-objective level of language (related to sender’s added commentary on the described events/situations and their participants) did not produce an immediate breakthrough in this respect, and a fuller picture of this group of expressions – rich and difficult to classify – continued to emerge over the years as an outcome of painstaking semantic-syntactic studies, with the majority of studies, out of necessity, tackling the problem of differentiation of the class in question from the expressions at the objective level. It is also in these studies that we see the terminology reserved for the description of the non-objective level emerging, and liberating scholars from the constraints of approaching adverbs as a collective category (see below, with special focus on studies by Maciej Grochowski). Furthermore, mostly thanks to the work of Jadwiga Wajszczuk (1997, 2005), scholars recognized the stratification of this level in itself – it was acknowledged that metapredicative expressions (adverbial comments) are placed at a lower level than metatextual expressions (added to an utterance as a sort of communicative whole). In consequence, the problem of re-definition of traditional adverbs in the context of the opposition of three, rather than two linguistic tiers, emerged, see item 2.

The basic semantic property of adverbs as a kind of superclass of second degree determinants highlighted in these studies, acknowledged also by Jerzy Kuryłowicz (1936/1979), Adam Heinz (1965), Renata Grzegorczykowa (1975) and other authors, is reflected in their traditionally identified syntactic functions. In a sentence, they typically perform the function of adverbials, i.e. either describe the circumstances (such as time and place) or qualitative properties of the main component (cf. this distinction in Kuryłowicz [1936/1979]), unlike complements which pertain to objects covered by a specific activity/property. The “objective” and “qualitative” moment of expressions in the context of discussions on parts of speech and syntactic functions was subtly described in particular by Heinz (1957) (see, however, also the controversy regarding the division of adverbs into qualitative and circumstantial, proposed

e.g. by Henryk Wróbel [1969: 19–20]). In connection with the evolution of adverbs, a similar distinction was applied by e.g. Grzegorczykowa (1957) as well as Zygmunt Rysiewicz (1956)

who, when commenting on the change of function of adverbial forms, identified the “syntactic revaluation of case” from the marker of relations between objects to “a more and more pronounced qualitative aspect” (p. 15), cf. e.g. obrócić się przodem / bokiem [with what?] → stać przodem / bokiem do kogoś [how?], pójść na przód czegoś [of what?] → pójść naprzód [where?].

Apart from identifying the qualitative/circumstantial nature of adverbs, some scholars also suggested (Tokarski 1967; Wróbel 1988) to equate the concept of a complement with the obligatory nature of complementation, while adverbials with the optional nature thereof, which, in the case of certain combinations, leads to recognizing the complementary function of ad-

verbs, e.g. czuć się / wyglądać dobrze, pachnieć / smakować wyśmienicie, brzmieć / zapowiadać się ciekawie, mieszkać / przebywać / odbywać się tam, wrócić stamtąd, traktować kogoś / obchodzić się z kimś / odnosić się do kogoś protekcjonalnie, postąpić / zachowywać się niedorzecznie, trwać długo (cf. more on such adverbs in the context of discussion on the opposition between arguments and modifiers, Chojak 2017). It was also recognized that adverbials may play a predicative role in a clause, taking over the role of the main segment of a sentence – both in the case of structures with a linking verb (Tu jest miło, Czytać jest przyjemnie) and without it (Miło tu, Duszno dziś). This function was recorded for instance by Zenon Klemensiewicz (1937), who nevertheless noted that not all uses with the verb być are predicative – in such sentences as Ona była późno, the verb has “the independent meaning of ‘being’” (1937: 106), while the adverb is an adverbial of place. Stanisław Jodłowski (1976) classified structures that typically have no linking verbs into uninflected verbs; see also the discussion of the class of predicatives below.

What is more, there is no consensus as to the role of adverbs in sentences such as Głośniej będzie lepiej, where traditional syntax would see them as the subject. The specification of this concept in the subsequent syntactic studies, i.e. restricting it to the noun in the nominative case imposing the person, number and gender on a personal form of the verb, mandates that we treat such structures as subject-free; the expression in the left-handed slot is rather a somewhat condensed representative of the situation thematized in the sentence, cf. Jeśli po-

wiesz to głośniej, będzie lepiej v. Lepiej będzie głośniej → Będzie lepiej, jeśli powiesz to głośniej. In view of such research, one cannot treat the following sentences quoted by Klemensiewicz

(1937: 101) as sample uses of adverbs in the function of the subject: Wczoraj było przyjemniejsze niż dziś, Owo „tam” życia przyszłego napawa nas niepokojem. In the first case, we are dealing with a noun homonymous to the adverb, as confirmed by the congruence with the adjective, while the other is a citation, implicitly embedded in the nominal structure. Structures such as Sumiennie to dokładnie (‘Sumiennie’ means ‘dokładnie’) could be interpreted as citations as well (cf. the discussion on “adverb in the role of subject” in: Brajerski 1995). Finally, even though the definition of an adverb as a property of an activity/property does not substantially envisage any combinations with nouns, adverbs have also been identified in the role of adnominal modifiers, cf. zakręt w lewo, drzwi na prawo, krok do tyłu, przyjaźń na wieki,

życie we dwójkę, życie na gorąco, jajka na miękko, kawa po turecku, chłopak na schwał. Some of these phrases form separate lexical wholes, but it is not a rule and many of them were created relying on a productive mechanism the use of which is not restricted to nominalization,

usually described as an exception to the identified selective restriction (czytanie na głos, spojrzenie ukradkiem, podróż zimą, rozłąka na zawsze, bieganie boso etc.). Researchers have approached this issue from a variety of perspectives, depending on the depth of interpretation: in her Polish-Russian comparative study, Emilia Wołodźko (1984) simply wrote that in such cases nouns are described by adverbs. Halina Koneczna (1933) spoke of adverbs in the function of adjectives, while Eugeniusz Cyran (1967) and Renata Grzegorczykowa (1975) considered such phrases as elliptical (‘jajko ugotowane na miękko’ etc.). In consequence, once again we

would be dealing with a case of the condensation of the initial structure (either occasional or lexicalized). As for combinations with nouns such as zgoła dziwak, prawie dziecko, bez mała pałac, na poły trup, niemal geniusz, they were usually disregarded in early grammatical compendia, though we find them e.g. in Adam Kryński (1897), who, relying on such examples, concluded that to define an adverb solely as a property of an activity/property is unjustified. By contrast, when analysing atypical adverbial collocations, Anatol Mirowicz (1947) found that in each of these cases we are dealing with a determination of a property anyway (here:

‘zgoła dziwaczny’, ‘na poły martwy’, ‘bez mała pałacowy’…); cf. the same position in Jan Safarewicz (1948), as well as Cyran (1967) and Grzegorczykowa (1975), who observed that nouns determined in this way may only perform a predicative function in a clause (“thus assuming the predicative function of a verb,” Safarewicz 1948: 50), cf. To zgoła dziwak. v. *Zgoła dziwak wszedł do pokoju.

On top of this, in all the abovementioned classifications of parts of speech – and, in fact, also the subsequent ones – morphological description plays a substantial role, as it allows for distinguishing adverbs from other types of expressions occurring secondarily in syntactic positions typical of adverbs (in line with the division into primary and secondary functions of words in a clause, Kuryłowicz 1936/1979). Obviously, the resistance to recognizing the adjectival nature of determiners in combinations such as sznycel po wiedeńsku has its origin in this approach. Except for the uninflected lexemes such as dziś, wczoraj, including “fossilized” specific cases (zimą, wieczorem…) and pro-adverbs (tu, gdzieś, kiedy, stamtąd…), adverbs are typically derived from adjectives using affixes -o or –e (originating from the Proto-Slavic forms of Nominative/Accusative and Locative, respectively, of neuter adjectives inflected according to the simple paradigm), with some of them being formed with the use of compound affixes such as po _-sku or po _-emu; see item 3. In grammars, this part of speech is substantially described as uninflected (invariable), with one major reservation that adverbs derived from gradable adjectives are also gradable – either synthetically (daleko – dalej – najdalej, including suppletive forms, e.g. dobrze – lepiej – najlepiej) or analytically (interesująco – bardziej interesująco – najbardziej interesująco). This forms a specific descriptive problem, sometimes producing some puzzlement, as in the “EJP” entry where we read that adverbs are “uninflected, but some [of them] are gradable” (1999: 312). In consequence, some scholars assume that gradation is not an inflectional, but rather a derivational phenomenon, cf. this approach in the section on word-formation in the so-called “academic grammar,” Grzegorczykowa 1998 (and the consistent omission of this category in the inflectional section of the study, Laskowski 1998b). Yet another solution consists in including such ad-

verbs into the adjective paradigm, for which gradation is a highly irregular category anyway. Importantly, and what substantially speaks in favour of the inflectional nature of gradation (cf. Tokarski 1973; Saloni 1974; Zaron 2003 et al.), is that adverbs derived from adjectives are usually gradable, and this is because the adverb-formation and gradation potential of adjectives are typically correlated with each other (though not without exceptions), meaning that both typically assume the “qualitativeness” of adjectives (as opposed to their “relationality,” cf. blady, zielony, ciekawy, uparty → blado, zielono, ciekawie, uparcie v. drewniany, szklany, podłogowy, żaglowy → *drewniano, *szklano, *podłogowo, *żaglowo; the possibility of deriving an adverb from an adjective is a proof of the “qualitatization” of the latter, cf. kolec

różany [thorn of a rose] → zapach różany [like that of a rose] → pachniało różanie, except for the role of limiting adverbials, such as grunty rolniczo nieopłacalne, where gradation is impossible; see again the “objective/qualitative moment” proposed by Adam Heinz, and, on the other hand, the arguments against this distinction, entry Adjective). In the inflectional model, descriptive gradation may either be classified into inflectional analytic structures or – as in the derivational model – to syntactically complex structures (cf. Tokarski 1973; Saloni, Świdziński 1981). What may support the latter solution is the fact that descriptive gradation may be both “positive” and “negative” (mniej_, najmniej_), and makes part of the general phenomenon of gradation/intensification, with highly diversified lexical markers, including the traditionally adverbial ones (strasznie nudnie, przeraźliwie blado, nieprawdopodobnie szybko, mało przydatnie, całkiem przyzwoicie, ciemno choć oko wykol, cicho jak makiem zasiał etc.); cf. the detailed works on this subject e.g.: Wierzbicka 1971; Janus 1975, 1981; Jurkowski

1976; Jurkowski et al. 1981; Szumińska 1997, 1998; Grochowski 2001, 2005, 2011; Chudyk 2006; Bałabaniak 2013; Linsztet 2014; Mitrenga 2014; Bałabaniak, Mitrenga 2015 et al. In both models – the inflectional and the derivational one – we further deal with the problem involving the differentiation between the units created as a result of inflection / semantically regular derivation from homonymous lexemes such as dalej [dziś pada] (*blisko / *najdalej [dziś pada]), najróżniej [określane] (różnie [określane] / *różniej [określane]) or lepiej [tam nie idź] (*dobrze / najlepiej [tam nie idź]), as well as – as in the case of gradation of adjectives – the issue of semantic differentiation of the comparative and superlative degree depending on the lexeme type, cf. Jan czekał dłużej od Piotra, choć obaj czekali krótko (relative property)

v. *Jan zachował się szlachetniej od Piotra, choć obaj zachowali się podle (absolute property) (cf. Grzegorczykowa 1972, 1998).

In consequence, relying on the early classifications of parts of speech which were mostly based on the semantic criterion and the related syntactic criterion, we arrive at the concept of adverbs as a multi-functional class, gathering lexemes capable of describing all independent

parts of speech with the exception of nouns, including adverbs themselves. This general capacity does not mean, however, that every unit of the class is endowed with it – it has been noted that only some of the lexemes are characterised by multi-functionality, while others are complementary to each other. In consequence, the detailed analyses of their syntactic distribution found in subsequent studies produced further distinctions within this class, and resulted in the identification of several new parts of speech. A side-effect of this precision was the need to consider certain expressions occurring in different types of contexts as homonymous. It is already in the classification proposed by Henryk Misz (1967) that we may find a strictly syntactic typology – on the one hand the author identified adverbs as dependents of verbs, but on the other he classified intensifiers as the dependents of adjectives and adverbs (sometimes homonymous with adverbs proper), cf. całkiem, zupełnie, dosyć, nadzwyczaj, niezmiernie, za, zbyt, zanadto, niemal, trochę etc. Likewise, in Roman Laskowski’s

(1984) syntactic classification found in the first edition of the GWJP, “Morfologia” volume, adverbs are described as autosyntagmatic components of a sentence, dependent on the main component; they are uninflected, performing the function of a higher level predicate or an argument expression connoted by verbs. Unlike nouns, adjectives and numerals, but similarly to “modalizers,” adverbs cannot make part of the nominal group; in contrast to the latter, however, they have restricted freedom of entering into syntactic relations. Lexemes which fill only adjective-adjacent and adverb-adjacent positions were thus excluded from this class (and included into the group of “modalizers”), which addresses previously raised doubts as to the equivalence of the status of adverbs in combinations with verbs and other parts of speech. It was already e.g. Safarewicz (1948) who noted that it is not correct to describe ad-

verb as a property of a property, as it would suggest that this is its primary function; yet the function considered primary should be typical of all words from a specific class, while only some adverbs collocate with adjectives and other adverbs. Władysław Cyran, in his monograph on word-formation history (1967), also viewed adverbs mostly as determiners of verbs. This approach was criticized in turn by Renata Grzegorczykowa, in the most comprehensive – to date – description of Polish expressions belonging to this class (1975). She emphasized that there is no other part of speech apart from adverbs that would have the function of describing adjectives and adverbs themselves.

As Jodłowski before him, Laskowski (1984) also excludes from this class expressions occurring in such combinations as Zimno mi / Przykro mu / Duszno tu / Mglisto dziś / Strasznie

parno, which were included in a subclass of verbs, i.e. (impersonal) predicatives (analytically inflected for tense and mood: Będzie mi zimno, Byłoby tu duszno, Było mglisto). This solution is supported – as noted by Solecka (1988) – by the possible repartition of endings, cf. Nudno tu – Mówił nudnie, Było mglisto – Wyrażał się mgliście, Było mu smutno – Uśmiechnęła się smutnie, Bardzo mi tęskno – Spojrzał na nią tęsknie etc. The remaining cases would be accounted for by homonymy (Wesoło tu – Uśmiechnął się wesoło). However, today the morphological distinction should be considered as residual, cf. Miło cię widzieć – uśmiechnęła się miło/ mile and Pochmurnie / pochmurno // skwarnie / skwarno dziś, Było tam gwarnie / gwarno, as well as Słonecznie / *słoneczno // upalnie / *upalno dziś, Zacisznie / *zaciszno tu. In accordance with Laskowski’s definition, predicatives should also include adverbial expressions in clauses with the so-called infinitival subject, where the infinitive is interpreted as the primarily right-handed complement of the impersonal verbal structure; cf., on the one hand, the order Czytać jest przyjemnie, repeated in grammars, and, on the other, Przyjemnie jest chodzić po plaży, Miło byłoby cię zobaczyć. Even though many grammarians argue that the predicative position as such should be considered as secondary, one should recall, quoting Stanisław Szober (1931), that it is the combination with infinitives that historically contributed to the morphological separation of adverbs from the class of adjectives (which, in this context, took the neuter nominative form).

Importantly, one should also emphasize that the predicative function of adverbs is not limited to the occurrences as given above, but involves every usage where the main stress in a sentence falls on this word (i.e. where it is the rheme of the sentence) or, potentially – as

in appositive structures – where we are dealing with parallel stress, resulting in a two-peak sentence stress structure, as already noted by Klemensiewicz (1937), as well as Krystyna Pisarkowa (1965) in the description of the so-called ancillary predicative phrases. The traditional syntactic function of a predicate is only one of its manifestations. Zuzanna Topolińska (1983) highlights this fact, within the apparatus of semantic syntax, by demonstrating that adverbs – just like adjectives – enter both restrictive and non-restrictive (appositive) combinations.

In the former, the adverb enriches the meaning of the verb, while narrowing down its reference (Przekupień głośno powtarza nazwy oferowanych przedmiotów). By contrast, in appositive attribution the reference is given (the activity has been identified), while the adverb only enriches its meaning (Przekupień powtarza nazwy oferowanych przedmiotów, głośno i zachęcająco). In certain cases, the actual carrier of predication is the adverb, while the verb may be even semantically empty (Jurek mówił to głośno, podczas gdy inni szeptali coś pod nosem). As shown above, we are dealing with a broader interpretation of predicativity than the one that refers to surface syntax relations. What is more, the possible occurrence of a lexeme in a predicative, rhematic position, becomes the fundamental test verifying its “actual” adverbialness, cf. e.g. Zachowywał się ZWYCZAJNIE v. Zwyczajnie mnie na to NIE STAĆ (*Nie stać mnie na to ZWYCZAJNIE); see item 2. In this approach, adverbs – in accordance with the differentiation introduced by Andrzej Bogusławski (1999) – are mostly primarily predicative/rhematic expressions (that may secondarily make part of the thematic dictum, i.e. the

presupposed part of a sentence), while metapredicative and metatextual expressions, as comments added to what is being said in a sentence are – by contrast – inherently thematic and cannot be used as rhemes. As Magdalena Danielewiczowa (2012: 150–162) noted, only certain types of adverbs typically (primarily) make part of the thematic portion of a clause: these include in particular limiting adverbials (Organizacyjnie wszystko było BEZ ZARZUTU, Merytorycznie nie mam NIC DO DODANIA; cf. more on this group also in Danielewiczowa 2021: 211–217), indexical markers of place/time and quantifying expressions such as rocznie, tygodniowo. Adverbs of degree, even though they usually make part of rheme, as broadly construed, do not typically take the main sentence stress within the rhematic dictum (Jest

niesłychanie PRACOWITY, Jest mi niezmiernie PRZYKRO), which is one of the reasons motivating the decision of some scholars to exclude them from the class of adverbs, see item 2. Nevertheless, deadjectival adverbs of this type (as well as temporal/locative expressions and, when it comes to the adverbs of degree, also bardzo) may, in specific contexts – unlike the inherently thematic ‘adverb-like’ expressions – occur in a rhematic position (Nasi słuchacze są już przygotowani IDEOLOGICZNIE, Przykro jest mi NIEZMIERNIE).

Researchers who consider the syntactic predicative function as derived include Grzegorczykowa (1975), as well as Henryk Wróbel (1988), who approaches the topic from the formal syntactic viewpoint and emphasizes that adverbs behave in this respect in a manner analogous to adjectives whose primary role is that of an attribute in a nominal phrase, cf. miła dziewczynka → dziewczynka jest miła and chodził niespokojnie → było niespokojnie; the author deems the status of lexemes in uses which typically lack any linking verbs as controversial. As far as combinations with infinitives are concerned, he assumes Saloni’s (1974) formal solution, according to which adverb is connoted by a defective third-person verb być (homonymous with the personal form). In this structure, the infinitive is an optional complement of the adverb (cf. Było mi tu przyjemnie – Było mi przyjemnie spotkać kolegów; Trudno o bilety – Trudno kupić bilety). As a result, the predicativity of adverbs is eliminated from the surface syntax. Unlike Laskowski (1984), but in line with Grzegorczykowa (1975), Wróbel classifies describing adjectives and adverbs as the defining functions of adverbs, along with describing verbs. In his subsequent study (1996), the researcher defines adverbs as non-independent lexemes, functioning as clause segments, non-connective (unlike relative pronouns and complementizers), which do not constitute clause heads and are not accommodated (i.e. do not accommodate any values of any grammatical categories with respect to the head they describe), additionally characterised by limited distribution, which sets them apart from particles (the scholar abandoned this view later on [2001]). In consequence, the latter are in direct opposition to adverbs, while “modalizers” are close to mood operators, and thus have a completely different scope than in Laskowski’s proposal. As a result, according to these authors, the “grey zone” between the adverb and the part of speech that is the closest to it

is located elsewhere.

The viewpoint presented by Zofia Zaron (1993) with regard to heads to which adverbs are attached and the secondary status of these expressions in predicative function is similar to the one espoused by Wróbel. Relying on the criterion of connotation and syntactic accommodation, as well as semantic and inflectional markers, the researcher describes ad-

verbs as lexemes without clause-forming function, which connote verbs, adjectives or other adverbs, are not accommodated and not inflected for case, number and gender, while being inflected for degree. Later on, in her functional classification of lexemes (2003), Zaron distinguished adverbs from localizers (elsewhere [2005] she explicitly writes about adverbs proper and adverbs-localizers): the former include lexemes derived from adjectives and bardzo (in compliance with Saloni’s [1974] morphological classification, where adverbs of this type were classified as adjectives, including also bardzo as synonymous with wielce [← wielki], and thus also więcej and najwięcej; see below), while the latter comprise expressions “localizing in time and space,” including pro-forms (dziś, jutro, wtedy, gdzieś…). The status of non-localizing adverbs that are not derived from adjectives is not clear (mimochodem, naprędce, znienacka,

po omacku, z kretesem…). Obviously, prepositional phrases may also serve as markers of location and – depending on interpretation – may be either considered as adverbial structures

(used to recognize the theme of an utterance by indicating its temporal/spatial location) or as analytical forms of a noun (provided one takes account of the objectivity [morphicity] of

the point of reference). This solution, suggesting the differentiation between the objective and qualitative aspect of a phrase already mentioned above, radically increases the set of syn-

tactic homonyms, which the author is ready to accept (Zaron 2005: 52).

The key contribution of Wróbel’s study (1988) involves the systemisation of formal-syntactic properties of adverbs – they are either elements connoted or not connoted by the head they connote. In consequence, the occurrence of an adverb determines the occurrence of its syntactic head, while the occurrence of a verb – depending on lexeme – either does or does not trigger the occurrence of an adverb; adverbs describing adjectives or other adverbs are non-obligatory dependents (except for situations when they are derived from expressions connoting adverbs, as in the phrase dobrze zbudowany). The following clause patterns have been distinguished: NP_ADVP (Jan wygląda młodo), NP_NP ADVP (Jan traktuje kogoś pogardliwie) as well as impersonal _NP ADVP (Janowi powodzi się źle) and _ADVP {(NP)(ADVP) (IP)(PP)(S)} (Piotrowi jest zimno chodzić bez płaszcza; Było mi przyjemnie, że spotkałem kolegów; Trudno było kupić bilety; the string {(), ()…} represents alternative connotation, i.e. any dependent except for one may be removed. In addition, adverbs may connote, and thus also accommodate, their own syntactic dependents (i.e. imply their occurrence and, at the same time, determine their grammatical form). This property is typical mostly of adverbs derived from adjectives with case government, e.g. [postępować] zależnie od czegoś, [zachowywać się] charakterystycznie dla kogoś, [w czymś jest] pełno czegoś, [zrobić coś] niezauważalnie dla kogoś etc., as well as comparatives and superlatives (wyżej / dalej niż_, najwyżej / najdalej z_). Nevertheless, the majority of adverbs take dependents optionally (bardzo dobrze, artystycznie słabo…), which – according to Wróbel – is typical i.a. of localizing expressions such as obok / blisko / niedaleko / nieopodal czegoś, daleko od czegoś, cf. a similar view in Węgiel (2000). By contrast, according to Piotr Wojdak (2004), independent occurrences of such lexemes are neither elliptical nor marked, and as such must have a separate status. Expressions without government are classified by the author as adverbs, while those with government – as prepositions. Cf. a similar view in a diachronic study by Wątor (1976), as well as Mirosław Bańko (2004: 112) who observes that the adverbs with case government require postprepositional forms of personal pronouns, and on this basis could be classified as prepositions (blisko niej, naprzeciwko niego, obok nich). In consequence, multi-segment prepositions are also distinguished, including w pobliżu czego, w poprzek czego (along with w czasie czego, w miejsce czego, na skutek czego, na bazie czego etc., even though they do not always pass the pro-form test). The homonymy of adverbs and prepositions becomes a necessary conclusion if one does not allow for the option that the former may attach dependents, cf. the class of non-connective

particles-adverbs in Saloni (1974), as well as the class of adverbs ‘proper’ as syntactically independent lexemes in Wajszczuk’s typology (2010). In her abovementioned study, Zofia Zaron (2005) starts with yet another set of assumptions, emphasizing that localizers always semantically connote the name of the point of reference, regardless of whether it is verbally indicated or not (in the latter case the point of reference is implied by the utterance itself). Therefore, at the semantic-syntactic level, these names are obligatory complements.

Inspired by the discussions regarding the heads of which adverbs are dependents, in his modified classification proposed in the 2^nd edition of “Morfologia” (Laskowski 1998a), based on the syntactic subordination relations and, additionally – in comparison to the 1^st edition – accommodation, Laskowski describes adverbs as dependents of a clause or nominal group,

which are not accommodated and do not combine with nouns (unlike “modalizers,” which may be combined with nouns [this also constitutes a change with regard to the 1^st edition] and which may include both sentence adverbials such as chyba, może, pewnie and – according to the author’s terminology – pragmatic operators such as tylko, zwłaszcza, szczególnie as well as quantifying words like prawie, zupełnie, całkiem, bez mała). Also, a separate group of predicatives, partially homonymous with adverbs, is distinguished, and both variants (1984, 1998a)

include a class of responses − contextual non-syntagmatic expressions (in contrast to non-contextual interjections) defined as lexemes that do not enter into syntactic combinations, consisting in short utterances that do not have the quality of a sentence; they are simply responses

to other utterances. This class comprises i.a. lexemes conformal with adverbs, such as istotnie, absolutnie, pewnie, dobrze, przeciwnie etc. (cf. [Zrobisz to?] -Dobrze / –Naturalnie v. Zrobił to dobrze / naturalnie), along with “adverb-like” expressions which do not have their equivalents among adverbs proper, e.g. oczywiście homonymous with the particle ([Pójdziesz tam?] -Oczy-

wiście v. Zrobił to oczywiście Janek); cf. Dobaczewski 1995, 1998, Wiśniewski 1995; for a different view see Wajszczuk 2005: 112 or Danielewiczowa 2012a: 32, according to whom distinguishing a separate “response” unit in such cases is tantamount to unnecessary multiplication of entities.

Subsequent descriptions proposed by grammarians and relying on semantic-syntactic relations introduced further, more precise distinctions with regard to expressions that may take other than adverbial positions; see item 2. On the other end of the spectrum, a classification based on the primary inflectional criterion was proposed (Saloni 1974; Saloni, Świdziński 1981). It took account of functional dependencies between elements of a sentence only to a minimum extent and when necessary, i.e. whenever morphological criterion has failed (i.e. with regard to uninflected parts of speech). In this approach, adverbs are traditionally divided into two inflectional classes – adverbs derived from adjectives are considered as the forms of relevant adjectival lexemes neutralized with regard to the case, number and gender (this group also includes adverbial forms of multiplicative, multiple and manifold numerals), while the remaining lexemes (including pro-adverbs) are classified as uninflected particles-adverbs, which differ from other non-independent parts of speech (conjunctions and prepositions) by their lack of connective function. Previously, the possibility of including adverbs derived from adjectives into the class of adjectives was examined by Heinz (1961),

who viewed this group as a borderline phenomenon between the (categorial) inflection and (irregular) derivation. Jan Tokarski (1973) was also explicitly in favour of this solution, which was rejected in turn by Grzegorczykowa (1975), who noted, among other things, that adjectives are not the derivational base in all adjective-adverb pairs, cf.: biegnie szybko → szybki biegacz, często choruje → częsta choroba, wykonać przed terminem → wykonać przedterminowo

→ przedterminowe wykonanie, półka położona wysoko → wysoka półka. In consequence, from the semantic perspective, one could distinguish between two classes of adjectives – the primary and the secondary ones, cf. ktoś jest miły → ktoś uśmiecha się miło → miły uśmiech. In the context of homonymy of adjectival forms, this issue was analysed by Andrzej Bogusławski (2005: 29), who saw adjectives with nominalizations as purely syntactic inflectional forms of adverbs; what makes the matter more complex, however, is the fact that the scholar is willing to interpret adverbs derived from primary adjectives in terms of inflectional categories as well, the difference lying in the possibility of semantic modification (cf. the example given by the author: Wszyscy w klasie są mądrzejsi od Zbyszka, ale Zbyszek tę sprawę ocenił najmądrzej ze wszystkich).

2. Adverbs in more specific semantic-semantic approaches; the problem of homonymy with other parts of speech

In his first “project of syntactic classification of Polish invariable lexemes,” Maciej Grochowski (1986) defines adverbs as uninflected (invariable) lexemes that are not used independently (unlike responses, onomatopoeia, predicative exclamations, appeals and interjections), and do not have the connective function (in contrast to prepositions and conjunctions, as well as relative pro-forms introducing subordinate clauses; the latter class was already proposed by Laskowski [1984, 1988a]). Adverbs may occur in declarative sentences (unlike mood operators and declarativity modifiers), but cannot be used as dependents of nouns, except for nomina actionis, which potentially retain syntactic properties of their derivational bases (cf. chodzenie szybko v. szybkie chodzenie). This last property distinguishes them from adnominal operators (bez mała, blisko, około, ponad…), particles (chyba, głównie, poniekąd, prawdopodobnie, przeważnie, widocznie…) and adnominal-adverbal operators (całkiem, niemal, odrobinę, po prostu…), which, in turn, differ from particles by their capacity to bind with proper names in the nominative case. (Given the exceptions with regard to the capacity to bind with nouns,

yet another researcher, Marek Wiśniewski [1995], did not include this property as a criterion differentiating particles from adverbs, replacing it with the criterion of connectivity with numerals, which adverbs – unlike particles – do not meet). According to Grochowski (1986), one distinctive feature of adverbs, apart from their syntactic dependence on a verb and/or adjective (the connectivity with adverbs in the context of definition is disregarded, to avoid the risk of a vicious circle), is the variable order which does not affect syntactic dependencies (and thus also the meaning of the clause), cf.: Jan na czczo poszedł na badania = Jan poszedł na czczo na badania = Jan poszedł na badania na czczo / Niebawem Jan dostanie pieniądze = Jan niebawem dostanie pieniądze = Jan dostanie niebawem pieniądze = Jan dostanie pieniądze niebawem (1986: 53–54), unlike in Chyba Jan przemawiał jako pierwszy ≠ Jan chyba przemawiał jako pierwszy / Głównie Ewa pomaga siostrze przy odrabianiu lekcji ≠ Ewa pomaga głównie siostrze przy odrabianiu lekcji ≠ Ewa pomaga siostrze głównie przy odrabianiu lekcji, where particles chyba and głównie bind with other (rhematic) elements occurring directly after them (1986: 64–65). It has been noted that the stress distinction of the rheme is indeed “more powerful” that the pre-location of the particle, cf. Ewa głównie pomaga siostrze / przy ODRABIANIU LEKCJI, where głównie applies to odrabianiu lekcji, even though it does not precede this phrase directly. Given this sharp distinction one should note, however, that the variable linear position is not typical of all expressions usually classified as adverbs: ‘subject-oriented’ adverbs (that describe both the subject and the activity, see item 4) pass this test with difficulty, while the markers of (non-)intentionality of an action, classified into this group, do not pass it at all, cf.: Jan przypadkowo WSYPAŁ sobie sól do herbaty (he did not want to put salt) v. Jan wsypał sobie sól przypadkowo do HERBATY (he wanted to put salt, but in his soup). Even if the position of an adverb is more or less fixed, its reference is determined by the scope of the rheme (which has effects with regard to truthfulness): Jan niechcący / przez pomyłkę wsypał sobie SÓL do herbaty (he wanted to put sugar rather than salt in his tea) v. Jan niechcący / przez pomyłkę wsypał SOBIE sól do herbaty (he wanted to put salt in Kasia’s cup); the first sentence implies that Jan did not want to put salt in his tea at all, while the other that he did want to put it in someone’s tea, just not in his own (see Duraj-Nowosielska 2021).

Eleven years later, Maciej Grochowski (1997) presented a modified classification of uninflected lexemes, taking account of the critical comments raised in response to his first classification attempt (e.g. by Wróbel [1995], regarding the application of non-syntactic category of proper name), and inconsistencies identified by the author himself. The key modifications concern the fusion of several subclasses of lexemes used independently and without context into a single class of interjections and the reorganisation of adverbial operators, for

which a new opposition was proposed: adnominal operators (byle [drobiazg], gdzieś [przed świętami], wprost [do garnka] etc.) were confronted with adnumerative ones (przeszło / z górą / ponad [pięć kilometrów], blisko / prawie / bez mała / niespełna [trzy miesiące], około / mniej więcej / plus minus [dwa tygodnie], równo [dwieście złotych] etc.). Some of these lexemes (such as wprost, gdzieś, blisko, równo) are homonymous with adverbs, as well as with adjectives (okrągły tysiąc, marne sto złotych, bite dwie godziny etc.) (cf. especially on adnumeratives Doboszyńska-Markie wicz 2013, 2020). The modification of criteria brought about a change in tensions within the system, although adverbs continue to be described as uninflected lexemes that do not function independently as utterances, that do not have the connective function, have a variable linear position and are incapable of entering into syntactic relations

with nouns other than the ones derived from verbs. Unlike Wróbel (1988, 1995), but similarly to Laskowski (1984), Grochowski excluded predicative expressions from this class (Nudno mi / Dziecku zimno / Ależ tu cicho! / Duszno dzisiaj), assuming that adverbs cannot be heads of a clause. In his work of 2001, the scholar once again undertook the issue of differentiation between adverbs and expressions generally referred to as functional, namely contextless particles (prawie, niemal), modal operators (oczywiście, podobno, właściwie, poniekąd etc.), adnumerative operators (bez mała, niespełna, z górą), adprepositional operators (tuż, wprost, zaraz) and gradation operators (bardzo, dość, nadzwyczaj, za, zbyt, zanadto; cf. the class of intensifiers in Henryk Misz [1967]). Once again, then, the set of metapredicates (with limited connectivity in comparison to particles and modal operators) was reorganised, and – in comparison to previous description – was extended by the class of gradation operators, traditionally identified with adverbs (of degree), cf. in particular the expressions with the form of adverbs derived from adjectives stosunkowo niewielki, względnie ogarnięty, relatywnie słaby, strasznie zarozumiały, kosmicznie bogaty, okropnie długi, wyjątkowo miły, zabójczo przystojny, masakrycznie nudny etc. (for more on emotive-intensifying expressions see e.g. Mitrenga 2014). As a result, scholars came across a specific problem of distinguishing this group from the group of adverbs ‘proper’, i.e. expressions from the objective level, cf.: Potraktowała go NIEMIŁOSIERNIE v. Jest niemiłosiernie GORĄCO; Spojrzał na nią wściekle v. Jest dziś wściekle

ZIMNO; Ujął to stosunkowo KOMPLETNIE v. Był kompletnie ROZTARGNIONY; Ujął to DOBRZE / NIEŹLE v. Był już dobrze / nieźle PIJANY; Ciągnęło się to NIESKOŃCZENIE v. Jest nieskończenie LEPSZY od niego; Zrób to dla mnie WYJĄTKOWO v. Był dziś wyjątkowo NIEMIŁY; Wypadł na tle klasy ŚREDNIO v. Jest średnio ZDOLNY; Napracował się już DOŚĆ v. Jest dość MIŁY.

The inclusion of gradation expressions as such into metapredicates may arouse some doubts, as many of such lexemes may combine with verbs and are capable of taking the main stress in a sentence, which is a property of adverbs, cf. Zachowywała się LEKKO i NATURALNIE v. Zupa była lekko CIEPŁA, Lekko się ZDZIWIŁA, with possible Zdziwiła się tylko LEKKO, on się zdziwił BARDZIEJ; Zachowywała się OKROPNIE v. Był okropnie GADATLIWY, Okropnie PRZYTYŁ, with possible Był gadatliwy / Przytył wprost OKROP-

NIE. As a matter of fact, the existence of the ‘grey zone’ between adverbs of degree and metapredicates should come as no surprise, as it reflects the evolutionary shift from expressions from the objective to the non-objective level, the final manifestation of which is the loosened or even broken link with the verb and inherent thematicity (*Był miły WZGLĘDNIE, as in *Nie był udany ZBYT etc.). When it comes to the possible occurrence of such lexemes in a rhematic position, Danielewiczowa (2012a: 157), unlike Grochowski (and also

e.g. Wajszczuk 2010: 65–66), perceives them as adverbs, even though primarily thematic in nature (cf. item 1). Their rhematicity (assuming a broad interpretation of rheme) is, in turn, unquestionable to Elżbieta Janus (1981). From this perspective, it is less of a problem to distinguish between adverbs and metatextual expressions (in Grochowski [2001] – particles and modal operators) which gain their actual meaning only in inherently thematic con-

texts, cf. the examples given by the author in subsequent studies: Przy nikim nie czuła się tak pewnie jak przy nim v. Jej adres jest już pewnie nieaktualny (*Jej adres jest nieaktualny PEWNIE); Z wiekiem zmienił się bardzo widocznie na twarzy – Widocznie było już późno, skoro zamykano wszystkie lokale (*Późno już było WIDOCZNIE); Postępował jak najbardziej właściwie – Właściwie nic się od tego czasu nie zmienił (*Nie zmienił się WŁAŚCIWIE). What is more, both the gradational and metatextual classes – just like the two remaining types of operators – contain expressions that are specific to them, including expressions having the form of adverbs derived from adjectives, cf. Przypuszczalnie nigdzie nie wyjedzie, Była nieporównanie sympatyczniejsza od siostry.

The subsequent years were characterised by intensive research on the non-objective level of the language, especially at the Toruń school led by Maciej Grochowski, resulting in further modifications. In his more recent works (Grochowski 2014, 2018, cf. also Grochowski, Kisiel, Żabowska 2014, 2018), the scholar writes about the three basic types of metapredicative expressions, traditionally referred to as adverbs or particles (occupying, however, a different place in the system of language than the ‘actual’ metatextual particles, including the modal operators mentioned above) – the gradation operators (which enter into one-sided relations with semantically gradable expressions, such as bardzo, całkiem, coraz, dużo, niewiele, wysoce), the approximation operators (mniej więcej, niespełna, prawie, niemal) and the limitation operators (indicating the limit of the property in comparison to the expectation of the speaker: aż, całkowicie, absolutnie, ledwo). As shown, the scopes of subsequent classifications overlap to a large extent, which renders their comparison even more difficult. According to the author himself (cf. e.g. Grochowski 2014), these groups are probably not exhaustive when it comes to the lists of metapredicates. Indeed, Magdalena Danielewiczowa (2010, 2012abc) argued that one should also include in this category the markers of attestation, construed as a kind of self-assessment of the sender with regard to the accuracy of the predicate used in the sentence, as e.g. Poruszaliśmy się w istnie/iście żółwim tempie; the majority of markers of attestation are homonymous with adverbs, cf. Był dosłownie wstrząśnięty v. Powiedział to dosłownie; Zostaliśmy regularnie oszukani v. Zajęcia odbywają się regularnie; Jego felietony miały charakter czysto informacyjny v. W domu było miło i czysto etc. (and with corresponding adjectives: istny cud, dosłowny absurd, regularne oszustwo, czysta bzdura, Danielewiczowa 2007). Due to their specificity, metapredicates do not combine with proper names, which distinguishes them from metatextual expressions, cf. (2012a: 117): Antykościelne hasła wykrzykiwał przypuszczalnie/oczywiście/naturalnie/niewątpliwie Janek v. *Antykościelne hasła wykrzykiwał jawnie/niespecjalnie/spokojnie/dosłownie/kompletnie/iście/czysto Janek (thus, the author relies on the criterion that Grochowski abandoned at some point). Nevertheless, they are characterised by a much freer connectivity than the operators of gradation, limitation and approximation. As a result, to set out the scope of this class in the context of other metaexpressions one needs to apply a number of detailed semantic-syntactic tests (cf. Danielewiczowa 2012a:

115–150). More on expressions of this type see also in studies by Dagmara Maryn (2010, 2019).

What is more, as shown by Danielewiczowa, in many cases, also the distinction between an adverb and a metapredicate is fluid. This applies in particular to the abovementioned adverbs simultaneously referred to an activity and the subject (and – one should add – contributing an element of assessment), e.g. lekkomyślnie, sprawiedliwie, egoistycznie, rozważnie, głupio, okrutnie etc., with regard to which one could observe an ongoing detachment from the objective level in statu nascendi, cf. the examples given by the author (2012a: 155) Sędzia wydał wyrok

ROZWAŻNIE  – Sędzia rozważnie WYDAŁ WYROK. In the thematic position they may even describe negative states of things, which is typical of metapredicative expressions: Sędzia rozważnie nie wydał wyroku. So sentence means: ‘The judge has not issued a judgement and it was prudent of him/her’, while “ordinary” adverbs imply that a specific situation has occurred – sentences such as Samochód szybko nie pojechał / Samochód nie pojechał szybko are interpreted as the negation of the adverb (the car did go, but not fast), while in specific context as the negation of both the verb and the adverb (the car did not go fast, because it did not go at all). The issue of expressions “attached to the subject,” only signalled in the study in question, which, according to the author, aspire to moving up to the higher tier of language, calls for further research, especially in the context of the impact of adverb’s position

in a clause on the scope of its reference, cf. e.g. Skonfrontował się z trudnościami odważnie /

śmiało (when confronted with difficulties he showed bravery/courage) v. Odważnie / śmiało skonfrontował się z trudnościami (just the mere fact that he decided to confront the difficulties was a sign of his courage/confidence [which does not determine courage/confidence during the confrontation itself]). This relates to the previously mentioned (although highly

variable depending on lexeme) syntactic “sensitivity” of “subject-oriented” adverbs. Such issues have been studied for a long time in English linguistics, a fact obviously motivated by the positional nature of the English language. Importantly, the changeability of reference in itself is not an indication to recognize polysemy, as emphasized by Bogusławski (2005), who proposed a general, monosemic semantic formula for the expressions of this type (based on the conjunction of predicates, see item 4).

Given the problem of homonymy outlined above, the studies by Danielewiczowa and Grochowski (as well as by other authors analysing the issue, cf. Wajszczuk 2010) dedicate a lot of space to tests that allow for distinguishing between metapredicative/metatextual expressions and expressions that belong with the objective plane. They include, in particular, the

test of non-contrastive stress, applied in the foregoing examples of opposite pairs, failed by metapredicates and particles (cf. Obecni jawnie głosowali PRZECIW UCHWALE v. Obecni JAWNIE głosowali przeciw uchwale; Chory dziwnie CHRAPIE [the fact that he snores draws attention] v. Chory DZIWNIE chrapie [he snores otherwise than in a manner that would not attract attention], Danielewiczowa 2012a: 82, 91). On top of this, scholars list the following test-verified properties of expressions belonging with the non-objective tier (Danielewiczowa 2012a: 81–92, as well as Grochowski 2008, 2014, 2018; Grochowski 2014 additionally organises the properties listed in the form of an algorithm): they occur in direct preposition with regard to the expression being described (Oskarżony bezwzględnie to powiedział v. Oskarżony powiedział to bezwzględnie); they cannot be negated (*Piotruś był niezdecydowanie szybszy v. Przegrał, bo działał niezdecydowanie; *W tym pudełku mieści się niespokojnie 20 kredek v. Motor pracował niespokojnie; niekoniecznie, niespecjalnie, nieszczególnie are independent metapredicates, cf. e.g. Danielewiczowa 2012b, Kisiel 2012); are not gradable (*Niektóre z tych cnót dziwniej wyglądają na cnoty stare v. Ten student zachowuje się dziwniej niż tamten; cf. separate expressions such as najzwyczajniej, najwyżej, najdalej, lepiej, najlepiej, prędzej, najprędzej, najwyraźniej in: Zwyczajnie/najzwyczajniej/*zwyczajniej się nie martw, Najwyżej/ *wysoko/*wyżej się spóźnisz, Oddam ci to najdalej/*daleko/*dalej za dwa tygodnie, Lepiej/ najlepiej/*dobrze zajmuj się sobą samym, Prędzej/*prędko się udławi niż przyzna komuś rację etc.) (Danielewiczowa 2012c; Grochowski 2020); they do not bind metapredicative operators (Zarabiał znacznie lepiej ode mnie v. *Kupmy znacznie lepiej pięć żarówek, a nie cztery; Odezwał się dość stanowczo – *To jest dość stanowczo za dużo). What argues in favour of distinguishing between the two separate classes is their possible co-occurrence in one sentence (Zamiast wyrzekać na innych niech on lepiej nauczy się traktować ludzi trochę lepiej; Postąpiłeś właściwie właściwie; Zmienił się na twarzy widocznie bardzo widocznie, skoro go nie poznałaś). Additionally, Grochowski (2008) lists the substantial non-transformability of a unit into the expression ‘w sposób_’ (Praktycznie nawet pracownicy nie byli tam wpuszczani − *W sposób praktyczny nawet pracownicy nie byli tam wpuszczani v. Postępował bardzo praktycznie – Postępował w sposób bardzo praktyczny), but this test has a limited scope of applicability due to selective restrictions of w sposób_ itself (cf. Danielewiczowa 2012a: 104–107).

Later on, Vyara Maldijeva (1995, cf. also Małdżiewa, Bałtowa 1995) also built upon the initial classification of uninflected lexemes put forward by Maciej Grochowski (1986). She proposed a major modification to the criteria and the final division, the key difference pertaining to the lexemes that cannot be used independently and do not have the connective function. Words from this group are divided on the basis of whether they may bind with a verb or not – the former subclass includes adverbials that attach to verbs and are governed by

them as well as particles proper that are not governed by verbs, non-indicative particles (which do not occur in declarative sentences), mood and tense operators as well as adverbal operators (distinguished on the basis of additional criteria). Lexemes that do not bind with verbs make up a further four classes of adverbs: adnominal adverbs not governed by adjectives (bezpośrednio, blisko, byle, dosyć, głęboko, jednakowo, miesięcznie, natychmiast etc.) as well as three classes of expressions governed by adjectives – ad-adjectival adverbs not governed by adverbs from other classes (bezwzględnie, całkowicie, czysto, dookoła, dotąd, nieporównanie, od wewnątrz,

przysłowiowo, wiecznie, wysoce, wszędzie, zwyczajnie etc.) and the ad-comparative (dwukrotnie, daleko, dużo, jeszcze, sporo, trochę, znacznie, nieco, niewiele, tym…) as well as ad-positive adverbs governed by them (nadzwyczaj, nadal, niesamowicie, niespecjalnie, ściśle, wyjątkowo,

względnie, wątpliwie, tymczasowo, ogólnie, osobliwie…); the list of adverbal adverbs comprises in turn such lexemes as bezpośrednio, celowo, dlaczego, dokądkolwiek, doprawdy, dosyć, inaczej, istotnie, mało, mimochodem, naprzód, niedaleko, niesamowicie, niespecjalnie, względnie, zasadniczo or zbytnio. It would seem that the degree of detail arising from the mechanically applied distributive criteria did not contribute to the consistency of description in this case. On the one hand, these sets are clearly heterogeneous, especially when it comes to the amalgamation of the objective and non-objective level (cf. e.g. doprawdy in the group of adverbal adverbs or czysto next to od wewnątrz in the ad-adjectival group, as well as tymczasowo with

ściśle or niespecjalnie in the ad-positive group). On the other hand, most adverbs have their representatives in at least two, and sometimes four or six classes (including homonymy with predicatives and responses). Eventually, this description is not particularly effective when it comes to the explanation and organisation of linguistic phenomena. Nevertheless, this does not preclude its usefulness for lexicographic purposes (as the author(s) emphasize(s)).

One classification of lexemes which – unlike Maldijewa’s approach, and in fact unlike any other prior study – starts by directly differentiating between the syntactic and metatextual levels is the proposal put forward by a scholar already mentioned above, Jadwiga Wajszczuk (2010; see also Wajszczuk 2005). In this case, the starting point is to demonstrate the distinctiveness of the sentence syntax units (or the elements of the syntax of dependency, or syntactemes) from the utterance syntax units (or the elements of the syntax of co-occurrence, or paratactemes). Both types are contrasted with paralexemes which, as single-class words, do not enter proportional orders typical of the linguistic systems (they include asyntagmatic units such as appeals or interjections, distinguished in earlier typologies). Syntactemes open semantically and grammatically marked positions and/or fill such positions in a sentence,

while paratactemes either do not have this property or they only co-occur with the units of the syntactic level, without imposing on them, as a class, any semantic or grammatical restrictions (these include particles opening a single position and – a substantive change in comparison to the approaches proposed by other authors – conjunctions that open two positions; cf.

Wajszczuk 1997). In this approach, adverbs are classified as syntactemes, and in this group are further described as independent words – i.e. words that do not open any syntactic positions themselves but fill positions opened by other lexemes. The author separately distinguished a class of predicatives, as well as the class of adverbial participles and metapredicative operators. In her study of 1997, the latter were construed as the elements of the metatext ‘proper’, but in 2005 the researcher abandoned this view. She continued to perceive them as metacomments overlaid on the objective level of an utterance, yet entering into ‘normal’ (although – in contrast to autosyntagmatics units – unilateral) syntactic dependencies with their heads (in consequence, they correspond to a specific transitional space between the objective and the purely metatextual level, which finds its reflection in the abovementioned controversies concerning their description). The comparative forms (of adjectives and adverbs) – once again, in an atypical way when confronted with other approaches – constitute a separate class of comparatives, which belong to a group of syntactically dependent lexemes, i.e. lexemes which both fill the positions opened by other words and that open such positions themselves. The status of the superlative is not entirely clear, but one might probably assume that

it also belongs with the “comparative” class (which opens a syntactic position for z/spośród_). Particles, distinguished at a different level as metatextual expressions, operate (just like conjunctions) in the communicative tier, or, in other words, on the thematic-rhematic structure of an utterance (which explains the free nature of their connectivity). As such, they cannot, in any case, be identified with adverbs, even though they may be homonymous with them; to be more specific, they pertain to the entire TRS, or to the rheme itself, being yet another criterion for the division. The opposition between adverbs and units of this type has already been discussed, (cf. e.g. Bogusławski 2009; Grochowski 2009, 2014, 2018; Danielewiczowa 2012; Grochowski, Kisiel, Żabowska 2014, 2018; Żabowska 2015 et al.

This overview demonstrates that the grammatical description of this class has evolved from observations on the multifunctionality of adverbs towards more precise definitions, resulting in expanding the set of lexemes homonymous with other parts of speech. The summary contains the examples of homonymy systematized from this perspective, with reservation that one can speak of homonymy only in the context of a specific classification. In consequence, this overview is merely indicative. Essentially the only part of speech for which no homonymous representatives in the group of adverbs proper were proposed is the class of adjectives, although the examples such as budynek po lewo analysed above could also in this case provide some grounds for this view (certain researchers speak of the adjectival function of adverbs precisely in this context). Owing to the relatively fresh discussion on the borderline between metaprediactive and metatextual expressions, only unequivocal specimens were included in the former. Applying a consistent research approach, it could be justified to identify the homonymy of units also in this area, cf. the following awkward but nevertheless seemingly non-tautological sentences: Stanowczo [I believe that:] wydał na to stanowczo za dużo, Zasadniczo [I believe that] wszystkie te urządzenia są zasadniczo do niczego, Absolutnie [I am of the opinion that:] jest na to absolutnie za młody (and the analogous: Stanowczo wydał na to zdecydowanie za dużo / Zdecydowanie wydał na to stanowczo za dużo, Absolutnie jest na to stanowczo za młody / Stanowczo jest na to absolutnie za młody).

Adverbs and nouns (in these examples, the first element of the pair indicates the use of an adverb): Zrobię to dziś / jutro – Dziś mija, jutro nadchodzi; Zrobiłam to wczoraj – Wczoraj

już nie wróci; Zrobiło się ciepło – Ogarnęło nas ciepło; Było zimno – Ręce zgrabiały od zimna; Pojadę tam zimą / latem  – Jestem zmęczona tą zimą / tym latem; Był z nią sam na sam – Podobało mu się takie sam na sam z nią (more on phrases such as od jutra, do rana, z prawa na lewo see in item 3). Adverbs and predicatives: Czuję się niedobrze – Niedobrze mi; Uśmiechała się miło – Miło mi cię widzieć; Przyszło mu to łatwo – Łatwo tak mówić; Wyglądała świetnie – Świetnie, że przyszłaś. Adverbs and interjections: Dość /dosyć się nacierpiał – Już dość / dosyć!; Przejdź naprzód – Naprzód!; Siedział cicho – Cicho!. Adverbs and responses: Zrobił to dobrze / Wygląda naturalnie / Stań tak – [Zrobisz to?] –Dobrze / –Naturalnie / –Tak; Właśnie przyszedł – [X nie przyszedł] –Właśnie; Rządził absolutnie – [Masz ochotę przyjść?] –Absolutnie. Różnili się od siebie bardzo istotnie – [Nie sądzisz, że powinniśmy już wyjść?] –Istotnie. Adverbs and numerals (pro-numerals): Trochę / niemało się natrudził – Zebrał trochę / niemało owoców; Dużo / mnóstwo o tym myślał – Przyszło dużo / mnóstwo ludzi; Ile on mówił! – Ile przyniosłeś książek? Niewiele się zmieniła – Ma niewiele zalet. Adverbs and prepositions: Mieszka blisko / niedaleko / nieopodal – Mieszka blisko / niedaleko / nieopodal dworca; Chodził dookoła / naokoło – Chodził dookoła / naokoło domu; Mieszka naprzeciw / naprzeciwko / obok / w pobliżu – Mieszka naprzeciw / naprzeciwko / obok / w pobliżu nas; Przekroił to wzdłuż – Biegł wzdłuż pociągu. Adverbs (pro-adverbs) and conjunctions: Zrób to jakkolwiek – W końcu by-

łam na czas, jakkolwiek wyszłam za późno; Wyjedźmy stąd – Nie uczyłem się, stąd nie dziwne, że nie zdałem; Kiedy przyjdziesz? – Kiedy ja (naprawdę) nie mogę przyjść; Jak mam to zrobić? – Jak zdecydujesz się przyjechać, to pogadamy. Adverbs (pro-adverbs) and relative pro-forms: Poszedłby za tobą dokądkolwiek – Dokądkolwiek pójdziesz, znajdę cię; Zrób to kiedykolwiek – Kiedykolwiek się zjawisz, chętnie cię przyjmę; Zrób to jakkolwiek – Jakkolwiek to zrobisz, będzie dobrze; Gdzie mam cię szukać? – Nie wiem, gdzie cię szukać; Jak mam to zrobić? – Nie wiem, jak to zrobić (on ambiguity of jak see e.g. Chojak 2005, 2009). Adverbs and metapredicative expressions (approximators, limitators, gradators): gdzieś za miedzą  – gdzieś (ze) dwa kilo; Podszedł blisko  – blisko trzy kilogramy; Znajdował się daleko – Był od niego daleko lepszy; Wyglądała w tym grubo – Wygląda na grubo młodszą od niego; Sporo się namęczył – Zarabiał od niego sporo więcej; Nacierpiał się dość / dosyć  – Miał tego dosyć / dość dużo; Zachowywał się

śmiesznie – Zarabiał śmiesznie mało; Powiększył się dość znacznie – Był od niego znacznie wyższy; Zrób to tak – Jest tak niski, że nie sięga do półek (on ambiguity of tak cf. e.g. Walusiak

2005); Mało mnie to obchodzi – Jest mało zdolny; Malował wybitnie – Jest wybitnie gadatliwy; Znudził mnie strasznie – Jest strasznie zarozumiały. Adverbs and other non-objective expressions (metatextual expressions and expressions of controversial status): Śpij spokojnie – Wydasz na to spokojnie (z) tysiąc złotych; Wygląda naturalnie – To jest naturalnie kłopot / Naturalnie możesz na mnie liczyć; Zachowywał się całkiem normalnie – Normalnie mnie wkurzasz; Zachowywał się wczoraj najzwyczajniej (w świecie) – On był najzwyczajniej (w świecie) głupi / Najzwyczajniej to zignoruj; Zrób to zwyczajnie – Zwyczajnie się nie martw; Rządził absolutnie – Absolutnie masz rację; Wyrażał się zasadniczo – Zasadniczo nie wie, co robi; Ułożyli sobie życie szczęśliwie – Szczęśliwie nie musieli tam jechać; Odmówił stanowczo – To stanowczo za mało / Stanowczo to zignoruj; Odmówił zdecydowanie – Zdecydowanie nie musisz się spieszyć; Powiedział to wprost – Był wprost ujmujący / Wprost nie mogę w to uwierzyć. Zaraz wracam – [To dureń] -Zaraz (tam) ‘dureń’; Zrobił to pewnie i bez wahania  – Pewnie chcesz tam iść / Zrobił to pewnie Jan; Zrobiła to właściwie – Właściwie nie chcę tam iść / Zrobił to właściwie dobrze; Właśnie przyszła – Zrobił to właśnie Jan; Zrobił to przypadkiem – Masz to przypadkiem przy sobie?; Robili to jednocześnie – Jednocześnie ci przypomnę, że masz u mnie dług; Ujął to ogólnie – Ogólnie nie mam nic przeciwko temu; Różnią się miedzy sobą bardzo istotnie − Istotnie, jest nieznośny.

3. Word-formation properties; the issue of multi-segment adverbs

According to Grzegorczykowa (1998: 524), 99% of contemporary adverbs are morphologically derived from their corresponding adjectives. This disregards the semantic dependencies analysed above, which may disrupt this proportion (due to the complication involved in establishing the correct direction of semantic derivation, in vocabulary studies the inverse derivational dependency is most often overlooked, cf. specifically on this issue Teresa Smółkowa [1977: 74] in her study examining the frequency/productivity of adverbial formations, using Witold Doroszewski’s Dictionary as the input material). Let us, however, recall the differentiation of the direction of derivation which, as such, according to this author (Grzegorczykowa 1975, 1998), testifies to the derivational (rather than inflectional) nature of adverbs derived from adjectives. Other arguments (1998: 525) include the fact that only some adjectives can be used to derive adverbs as well as the fact that one adjective may have several semantically diversified equivalents, cf. wysoki − wysoko – wysoce, pański – pańsko – po pań-

sku, suchy – sucho – na sucho – do sucha. Adverbial forms cannot be derived from: 1. the majority of adjectives in the so-called relational uses (with the “objective moment” exposed, cf. above), indicating the relation between the referent of the noun and the referent of the nominal base of the adjective, even though the same adjectives (in certain approaches: adjectives homonymous with them) in qualitative use are not subject to this restriction (cf. koński ogon

− ogon konia − *machnął końsko ogonem v. koński uśmiech – uśmiech jak konia – uśmiechnął się końsko); relational adjectives of certain type may only be used to derive limiting adverbs (uniwersytecko obyty, mieszkaniowo dobrze ustawiony etc.); 2. certain adjectives derived from verbs that retained the case government, e.g. skłonny / zdolny do czegoś (*zdolnie / skłonnie do czegoś); 3. adjectives derived from adverbs dzisiejszy, jutrzejszy (*dzisiejszo, *jutrzejszo); 4. ordinal numerals pierwszy, drugi… (*pierwszo, *drugo…); 5. adjectives that occur only in predicative function (winny, rad). The repartition of -o/-e affixes is influenced by phonetic and derivational factors (cf. Śmiech 1957; Wróbel 1966; Cyran 1967; Grzegorczykowa 1998). What merits particular attention in this context are double adverbs (kwiecisto – kwieciście, wyrazisto – wyraziście, promienisto – promieniście), especially in the case of explicit repartition of meaning, as in Pokroił to równo – Była równie miła, Zaszedł wysoko – Był wysoce niemiły, Mieszkał daleko – Był daleko / dalece lepszy (apart from the practically already erased opposition between predicate and attribute, discussed above). This derivation type is highly productive, especially in the variant with -o (cf. internetowo, mainstreamowo etc.), and thus – despite the identified irregularities – it is sometimes proposed to include adverbial forms into the adjective paradigm (cf. above), while in operational grammar (Bogusławski 1978), the presence of the relevant adverbial operation has been postulated (in semantically justified cases), see Bogusławski 2005. Other productive formations, including the borderline cases between lexis and grammar, may be approached in an analogous way, cf. the same author (2005) on structures with a “prosthetic” noun, such as zrobić coś [jakimś] ruchem / gestem, iść [jakimś] krokiem, powiedzieć coś [jakimś] tonem, spojrzeć [jakimś] okiem na coś or Heliasz,

Wójcicka (2012) on such structures as _litrami, kilometrami, (całymi) godzinami; see a discussion on the status of such expressions in: Kubicka 2014. Meanwhile, on the opposite pole, there are non-derived temporal and locative expressions (mostly pro-forms), as well as lexicalized specific cases with broken link to the derivational nominal base, e.g. words formally in the instrumental case, such as okrakiem, cichaczem, pędem, żywcem, ciurkiem, chyłkiem or consisting of two segments such as po omacku, w poprzek, na przekór, do szczętu, do cna, na wznak, bez ustanku, z kretesem, na wylot, na schwał, na glanc, na wskroś, na chybcika, bez liku etc. From the word-formation perspective, still transparent to a certain extent, but no longer productive, are the compounds such as wtenczas, pojutrze, naprzód and the instrumental and prepositional formations as wieczorem, zimą, kłusem, galopem, bokiem, tyłem, półgłosem, na oślep, na pewniaka, na pozór, na przemian, do reszty, po kolei, za bezcen etc.

Apart from the paradigmatic -o/-e type, yet another currently productive type is the prefixation-suffixation type po _-u, with adjectival bases ending in -ski (-cki), derived from adjectives that were in turn derived themselves from the names of countries, regions and cit-

ies, as well as names for persons, cf. po angielsku, po warszawsku, po mistrzowsku, po koleżeńsku, po ekspercku etc.; when it comes to the adjectives with other affixes, similar structures are created with the use of po _-emu, e.g. po chłopięcemu, po młodzieńczemu, po psiemu. With several exceptions, such formations are not included in the overview made by Smółkowa (1977); that is because they are not recorded in the base dictionary, even though one may find there such words as kołnierzowato and bramiasto − and this is most likely the consequence of the fact that the author of the dictionary classified them as syntactic combinations; currently the WSJP lists them in separate entries. Apart from established collocations, such as mówić po polsku, Wróbel (1966) and other researchers mostly point out to the double – noun and adjective-based – derivational motivation (as a result, the affix would have the form of po _-sku/-cku) and the comparative meaning that these expressions entail, in contrast to derivatives ending with -o, cf.: Fakty układa po kronikarsku ‘like a chronicler’/ Zachowuje się po chamsku ‘like a boor’ v. Kronikarsko ujęte wydarzenia ‘in a chronicle style, concisely’, Zachowywał się chamsko ‘in a boorish way’. Yet another semantic modification is introduced by the type z _-a derivatives, which denote the weakening of a property, cf. nosić się po pańsku / po francusku ‘like a lord/ Frenchman’ v. nosić się z pańska / z francuska ‘somewhat like a lord/Frenchman’. The formations derived in an analogous manner from adverbs, such as z cicha, z lekka, z dawna, z wolna, also form a group which is somewhat akin to diminutives. Since adverbs as such are not accommodated, currently the strings of this type are considered as lexical derivatives formed in accordance with a specific model, cf. na gorąco, na sucho, (jajko) na twardo, na miękko; do sucha, do naga, do czysta; z bliska, z dala, z wysoka etc. (and yet, e.g. Tokarski 1949, suggested to perceive them as isolated inflectional forms of adverbs).

Meanwhile, combinations with prepositions, such as od jutra, do rana, z rana, given the identified nominal use of these lexemes (similarly przed południem, po południu, pod wieczór etc.) are typically interpreted as “active” prepositional phrases (in this case the preposition synchronically imposes the case on the noun). Nevertheless, not all of the analogous formations may be interpreted as prepositional-nominal structures, i.e. it is substantially impossible with analogous uninflected words, such as na/od/do kiedy/teraz/zawsze/wczoraj; they must thus be classified as short series of lexical derivatives. According to certain authors, e.g. Monika Czerepowicka (2006), the lexicalization of two-segment combinations does not preclude the identification of their “internal” grammatical properties, i.e. it is assumed that lexicalized adverbs, just like other phrases, may have their own internal syntax (cf. Lewicki 1986). The approach to phraseology declared by the author results in the recognition of separate units such as omacek, oścież, wskroś, wspak, jaw, przekór, which nevertheless require a dictionary note on their extremely limited lexical connectivity. The concept of a bilateral unit of a language in the meaning given to this term by Bogusławski (1976) does not allow for such a solution, as the components of these structures do not enter with one another into mutual proportional structures. Piotr Wojdak (2004) also adopts a purely formal view, proposing a systematized description of such combinations relying on the generative syntax model, cf. e.g. the following strings: bez ogródek / osłonek / przesady / miary; do

cna / dna / luftu / szczętu; na oklep / oko / opak and many others, including adverbial multi-segment combinations, e.g. with jak (jak z bicza strzelił, jak w mordę strzelił, jak makiem zasiał…). See also the analogous approach represented by Katarzyna Węgrzynek (2005) in the description of structural strings such as iść na parking – iść na emeryturę − iść na noże or in her analysis of repetitive structures (2000) ramię w ramię, oko w oko, etc. By contrast, the authors of “Słownik reduplikacji” (Dobaczewski et al. 2018) adopt a consistently lexical approach to the second group of such expressions, cf. adverbial wholes ręka w rękę, słowo

w słowo, od słowa do słowa, krok po kroku, noga za nogą, kubek w kubek and other, in contrast to structures creating syntactic series such as dzień w dzień, od domu do domu, z tygodnia na tydzień, dzień po dniu etc.; cf. the discussion on unit/serial nature of such structure in: Rosalska 2011; Gębka-Wolak, Moroz 2014; Dobaczewski 2018.

4. Semantic types of adverbs

When it comes to the early proposals of a relatively systematic approach to this class in terms of semantic differentiation, we have already mentioned the various types distinguished by Henryk Gaertn er (1938) which, when it comes to the degree of their specificity, have not been matched neither by the scholar’s contemporaries nor in the subsequent studies. Given the semantic nature of the project, this actually comes as no surprise. A more detailed classification was only presented by Janina Nowakowska with regard to adverbs of manner; the scholar dedicated a separate monograph (1933) to this group. The most comprehensive attempt at the semantic organisation of adverbs may be found in the study by Renata Grzegorczykowa (1975), already quoted multiple times in this chapter. There we will find a number of topics discussed herein, including the identification of a class of “modifiers of entire clauses having the modal nature or markers of condensation of several clauses” (p. 26). These are, therefore, the equivalents of what was subsequently described as metatextual expressions, of which the author explicitly writes that “they actually do not belong with the class of adverbs.” In the

remaining cases, the proposed main axis of division is the type of predication, going back to the logical-semantic properties of expressions, which refers to the concept put forward by Hans Reichenbach (1967). Therefore, the predicates of the first degree describe objects in the function of subjects (X umarł bezdzietnie/młodo = ‘X died and X was [at that time] childless/

young’, X wygląda młodo ‘X looks young’) or complements X ożenił się bogato “X married someone who is rich,” X pomalował ścianę na biało “X made the wall white”). Apart from the qualitative description of objects, adverbs of this type may convey their quantitative description as well (Goście przybyli licznie = “The guests came and the guests were numerous”). As expressions which do not refer to actions/properties, they are actually only apparent adverbs, which narrows down the class being described even further. The set of adverbs proper consists only of the second degree predicates, referring to other predicates, i.e. in accordance

with the adopted assumptions – verbs, adjectives or adverbs (in logical-semantic terminology, we are then dealing here with the class of modifiers rather than predicates). Adverbs which describe both an object and activity form a transitory class, as in sentences Odezwał się złośliwie, Postępuje sprytnie, Przygląda się podejrzliwie, Spojrzał na nią wrogo, Podskoczył wesoło (cf. ‘subject-oriented’ adverbs described above); what is of key importance in such combinations is that the verb being described names an activity/state in which a given property manifests itself (cf. siedzi smutnie v. *zna prawdę smutnie). The first degree predicates further include situation descriptors, including the descriptors of natural phenomena, cf. such already discussed structures as Jest tu odludnie / przytulnie, Zrobiło się wilgotno / ciemno, Wokół jest

biało od śniegu, Brudno tu, Przykro mi to mówić, Do domu jest już niedaleko etc. Even though some of the adverbs occur only in this role, the predicative function is substantially treated as secondary in this class, cf. e.g. the predicative occurrences in such structures as postawić

coś blisko, rozłożyć coś luźno, powiedzieć coś cicho etc. (pp. 32–33); as can be seen, the concept of predicativity in this case is inclusive of the communicative structure of an utterance, going beyond the prevailing understanding of the combination of a linking verb and a predicate in the grammars of that period. The division into modifiers and predicates is also known from the subsequent Polish studies on adverbials, cf. in particular Ozga (2011) below, as well as Bogusławski (2005), who justifies in detail the predicative nature of adverbs derived from primary adjectives (describing individuals, see item 1), proposing for them a general formula in terms of conjunction of predicates (i.e. double assertion) and their bilateral implication (cf. an analogous approach with regard to adverbial participles Weiss 1977); according to the scholar, this view is supported – among other things – by the impossibility of iteration inherent to the implication systems, as in *On niemądrze gniewnie spojrzał na nią. The study is theoretical rather than material-based, and thus it is difficult to unequivocally conclude

which specific adverbs would, according to the author, fall into this group, and what their approximate percentage share in the overall class of adverbs would be. Relying on the exam-

ples and comments (especially criticism with regard to the modifying approaches available in literature), one may conclude that the author would be ready to extend this definition to a much broader group of expressions than the one resulting from Grzegorczykowa’s work. First of all, this class would most likely include adverbs describing the subject and the activity simultaneously, which raises the question of whether the conjunctive approach is indeed sufficient. As regards subsequent studies inspired by that article, where researchers rely on the operation deriving adverbs from adjectives, one could also conclude that such operations were interpreted very broadly, as a relatively general mechanism applicable to deadjectival adverbs, while it seems to refer to only one of the semantic subclasses (see the distinctions below). By contrast, Zofia Zaron (1993) represents a completely different view. For her, the discussed differentiation has no justification in language, as words described as adverbs of the first degree modify the verb in the same way as those of the second degree. This rigid modification approach seems to originate from the programmatic identification of the surface-component structure with the semantic structure.

Grzegorczykowa (1975) subordinates her detailed semantic differentiations to the division of adverbials into determiners of adjectives and verbs; naturally, therefore, adverbs end up in two different syntactic classes. As regards the former, it is noted that only certain semantic types of adjectives may be described by adverbs and only certain types of adverbs may describe adjectives. In addition, a preliminary distribution of adverbs from this group from the perspective of semantic classes of adjectives is outlined (p. 45). These types of expressions are subdivided into: (a) adverbs of degree, conveying the intensity of a property, e.g. lekko

mokry, mocno niebieski, mało wyraźny, słabo widoczny; (b) qualitative adverbs, specifying a property, e.g. papierowo blady, dziecinnie naiwny, rdzennie polski; (c) limiting adverbs. ga-

tunkowo lichy, objętościowo niewielki; (d) adverbs referring to (the capacity of) performing or being subjected to an activity, e.g. trudno dostępny, łatwo wykrywalny, zmysłowo poznawalny etc. Adverbs of degree (a) are further divided into: (aa) adverbs conveying intensity, pointing out to (aaa) a very high degree of intensity of a property (bardzo mądry, bezdennie głupi, nieziemsko piękny), (aab), an average degree (średnio wysoki, w miarę dopasowany, przeciętnie zdolny, względnie suchy), (aac) a low degree of intensity (lekko niebieski, mało rozgarnięty, trochę ciepły); (ab) conveying information about the (in)sufficiency of a property in comparison to a norm, i.e. (aba) when the degree is too high (za gruby, zanadto rozległy), (abb) sufficient (dość ciekawy, dosyć gęsty [today such expressions in a non-stressed position would be rather interpreted as [aab]) and (abc) insufficient (za mało otwarty, nie dość ciepły); (ac) markers of (in) completeness of a property, describing adjectives that convey polar (non-gradable) properties and indicating (aca) a complete degree (zupełnie nagi, całkiem pusty) or (acb) an incomplete degree (prawie pusty, niemal doskonały); this division overlaps with the markers of gradation, approximation and limitation in Grochowski (2014).

Grzegorczykowa divided the descriptors of verbs into quantitative, qualitative, limiting, and circumstantial. The first group includes: markers of general and partial quantification of the predicate (na ogół, przeważnie, nigdy…), expressions conveying frequency (dwukrotnie, rzadko, jednorazowo as well as constructions [ile] razy, co tydzień / godzinę, za [którym] razem, po raz [który]…), duration ([coś trwa] jak długo / od kiedy, including structures such as godzinami, całymi dniami), and finally expressions conveying the “measurement and the degree of intensity of a state” (p. 48) (dużo [czyta], całkiem [zmókł], dziko [się niepokoi]…); by analogy to the adverbs derived from adjectives, subtypes are considered among the latter in terms of degree of intensity, completeness and sufficiency, also in the context of the semantic types of the verb (p. 71). A separate, quasi-adverbial group is made of expressions that quantify ob-

jects, cf. Poziomki rosną obficie, Fakty opisywano wyrywkowo, Pracują nad tym zespołowo, Wychodzili kolejno, Przyszli razem, Zrobił to w pojedynkę. When it comes to qualifying adverbs, apart from the sample specimens of apparent adverbials describing objects and borderline adverbials that describe both objects and activities/states, the author includes in the group the adverbs of manner (pisać ręcznie, drukować maszynowo, ujmować rozumowo, zbadać doświadczalnie…), speed (powoli, jednostajnie, nagle…), adverbs that separately determine the physical properties of an activity (lekko potrącił, pali się słabo, that cannot be rephrased using the structure *w sposób lekki/słaby), including a special type describing an activity by means of the acoustic properties that accompany it (stąpa głośno, wszedł po cichu), conveying a degree of difficulty (mozolnie układał, dojechał łatwo), and finally entailing the assessment in

terms of: efficiency (szukał daremnie, pracuje na próżno), aesthetics (pięknie deklamuje, ładnie tańczy), morality (postępuje niegodziwie), as well as purpose (ołówek pisze dobrze) and the optimum state (znakomicie gotuje, kiepsko mówi). As a separate type, the scholar distinguished adverbs which introduce secondary predication, e.g. X bije kolegów bezkarnie = ‘X beats his

peers and it is not being punished’, Wystawiono sztukę próbnie, Chodzi śpiewając, including such structures as Śpi na siedząco. In certain cases, adverbs do not convey information about the auxiliary, but rather the basic activity, while the verb is semantically empty, or potentially contributes causative meaning, cf. działa usypiająco, kojąco (in this case, the subject does not perform two activities, “acting” and e.g. “calming down,” but only one of them – the one communicated by the adverbial, p. 98). In fact, qualifying adverbs differ from the apparent ones in that they do not convert into adjectives describing objects, e.g. X pisał ręcznie − *X jest ręczny, in contrast to X mówi naiwnie – X jest naiwny; the descriptions such as szybki biegacz point out to adjectives derived from adverbs (biegnie szybko). In consequence, different mechanisms behind the adverb-adjective pairs are thus revealed, and they are far from the automatism of morphological derivation, cf. such already mentioned examples as X jest świadomy – X zrobił coś świadomie – To był świadomy czyn X-a.

With respect to the limiting adverbs binding to verbs, the scholar concluded that certain expressions occur exclusively in this role, while other obtain this meaning in a specified context (though, obviously, this possibility does not apply to all adverbs in general), cf. Wyrażenia te różnią się znaczeniowo, Utwory dzielą się tematycznie na kilka grup, Nie wytrzymuje kondycyjnie. And finally, circumstantial adverbs include the adverbs that locate a situation in time and in space and inform about the relation between an event and other specified events. The group of temporal adverbs is divided into the deictic series (dziś, jutro, przedwczoraj as well as such structures as za dwa dni, tydzień temu), anaphoric series, including the precisely specifying expressions (wtedy, w tym czasie, miesiąc wcześniej, nazajutrz…), generalizing ones (kiedy indziej, potem, następnie, poprzednio…) as well as normative (wcześnie,

późno). In addition, some expressions of complex semantic structure containing a temporal element have been distinguished, such as od razu or natychmiast, as well as już, jeszcze, do-

piero (cf. for a different view e.g. Bańkowski 1976, who perceived this final class as particles). The group of prepositional locative determiners consists of expressions with specific quantification (tu, tam), specific indefinite quantification (gdzieś), specific indefinite partial quantification (gdzieniegdzie), specific indefinite quantification excluding a specified place (gdzie indziej), specific free quantification (gdziekolwiek), general positive quantification (wszędzie), general negative quantification (nigdzie) as well as interrogative expressions (gdzie). These types (although not all of them) have their ablative (skądkolwiek, zewsząd…), adlative (dokądś, donikąd…) and perlative equivalents (tamtędy, którędyś…). For more on prepositions and the locative pro-forms related to them cf. Klebanowska 1971, Weinsberg 1973, and from more recent studies Alberski 2018, 2019. In terms of the third type of circumstantial relations, it is emphasized that it is only exceptionally that they are expressed purely lexically, cf. Pracuje zarobkowo → he works to make money (purpose), Przeniesiono go karnie → he was transferred, because he was punished (cause); this type is similar to the auxiliary predication identified in another place in this study. Nevertheless, purpose, condition, cause and the lack of condition (concession) are typically expressed using syntactic structures.

The above description, which provides a good overview of the very numerous and highly diversified class of adverbs, despite obvious efforts of the scholar to ensure the transparency of the classification, demonstrates what a hard task it is to attempt a panoramic review of this part of speech. Certain types seem to overlap, and as the properties of many lexemes may only be described in their syntactic environment, in the context of the distinguished classes we are faced with the question regarding the status of variants (a similar phenomenon, to an even greater scale, may be observed with regard to adjectives). What is more, the different types of adverbs have been described applying varying degree of precision. On top of this, the described classes – contrary to the title of the study, but in line with the author’s goal to arrive at a comprehensive description – include, apart from adverbial lexemes, nominal and numeral structures that perform the function of adverbial phrases. Furthermore, the detailed description fails to provide clarity as to the distinction, set out initially, between predicative and modifying expressions that occur in the various parts of the analysis, and thus fails to

meet an important criterion role.

Four decades later, Kazimierz Ozga (2011) attempted to build his description by relying consistently on the distinction of this type. However, his focus was not only on the Polish language, and the scope of Polish material was incomparably smaller in comparison to Grzegorczykowa’s work. In his comparative analysis of sample Polish, Russian and English adverbials (using the apparatus of communicative grammar enriched with the selected concepts borrowed from semantic syntax), the author assumed that only adverbial predicates of the second degree have the isomorphic syntactic and semantic structure (i.e. adverb as a separate predicate indeed does describe/modify semantically the actions/states expressed by the verb, in compliance with the surface formalisation). In structures where an adverb performs the function of a predicate referring to the object, as well as in intensifying and temporal-locative structures, the semantic structure does not correspond to the formal relations; the correspondence or the lack thereof may be only determined by way of an in-depth semantic analysis (as earlier attempted by Grzegorczykowa), approached in Ozga’s proposal in a formalized way. In consequence, intensifying expressions do not introduce any exponents of separate concepts, but rather point out to the intensity of referents of higher-rank expressions,

which corresponds to the practice of accounting for lexical and grammatical gradation in a uniform way (see item 1); as regards earlier studies, cf. the view presented by Cyran (1967: 12): “adverbs of degree do not have any independent semantic value in themselves and they are used only to modify the meaning […], which approximates their role to the one of prefixes and suffixes.” In Ozga’s study, some of such expressions (zarąbiście dobry film etc.) further

include an expressive element (they are operators of interaction) and as such imply the direct orientation toward the sender, while others entail in their semantic structure a reference to a norm (za wysoki stół). Likewise, locative and temporal markers are also treated as special types of operators (so-called ideational ones). Markers of separate predication (always of non-isomorphic structure), in turn, are subdivided into several types, including e.g.: (a) with implicit subject identical with the subject of the verb, further subdivided into the following subtypes: Siedzi nieruchomo, Chodzi boso implying a conjunction (‘he sits/walks and is still/ barefoot’); Odruchowo sięgnął do kieszeni, Spoglądał badawczo, Ryknął gniewnie, with a causative relation between the adverbial predicate and the main predicate (the author notes the difference between the conjunctive zgodził się milcząc ‘he consented and was silent’ and the causative zgodził się milcząco ‘his silence was a sign of consent’ – earlier, Grzegorczykowa identified a similar distinction, but without deducing from it any systemic dependence as to the type of predicate; Uśmiechnęła się prowokacyjnie, which illustrates an opposite relation, i.e. the relation of effect (the smile elicited [or rather one should say: was supposed to elicit] a provocation); (b) with implied subject other than the subject of the verb – this type introduces the conceptualizer, e.g. Wrzasnął niespodziewanie / Wymknął się niepostrzeżenie

(for someone); (c) with an independent “scenario” introduced: Został pośmiertnie odznaczony,

Na próżno szukali broni (cf. adverbs signifying ancillary predication in Grzegorczykowa); (d) quantifying objects, e.g. Zgłaszali się kolejno. Moreover, there are also semi-isomorphic structures which include e.g. adverbials that fill argument positions. Despite the proposed formal tools, which encourage analytical precision, the preliminary nature of this description, given the immense richness of adverbs in terms of their number and the diversity of semantic structures, makes it difficult to assess the effectiveness of the distinctions introduced in the context of more vast linguistic material. It is probably due to the compounding difficulties faced by researchers attempting a comprehensive overview of this class of expressions that − save for the studies discussed above – in Polish linguistics there are no other monographs of equal theoretical ambitions. Rather, in their semantic analyses, researchers focus on a precise description of specific adverbial and adverb-like semantic fields (see e.g. Nowakowska 1933; Janus 1981; Linde-Usiekniewicz 2000; Chudyk 2006; Maryn 2009; Duszkin 2010; Heliasz 2012; Danielewiczowa 2012a; Bałabaniak 2013; Doboszyńska-Markiewicz 2013; Kubicka 2015; Duraj-Nowosielska 2021; Rosalska 2022).