Časopis Slovo a slovesnost
en cz

On the word order of Polish synsyntagmatic lexemes

Maciej Grochowski



Ke vztahu pořádku slov a synsyntagmatických lexémů v polštině

1. The paper presents the question of how it is possible to include the word order criterion into syntactic classification of lexical units. Below, the results of the author’s investigations on the linear order of subset of Polish lexemes, namely the synsyntagmatic ones, are discussed.

It is assumed that only a classification in accordance with homogeneous criteria, and only a classification subsequent to the logical divisions ought to be allowed. In the last decades numerous methodologically adequate grammatical classifications of Polish lexemes were proposed (cf. Saloni, 1974; Laskowski, 1984, 1998; Wróbel, 1996; Grochowski, 1986, 1997). Nonetheless, some of the widespread criteria of divisions should be regarded as controversial. So it is difficult to state if given units really fulfil the conditions which must be met as classification criteria. We very often hesitate whether or not to include a given lexeme into a given grammatical class earlier distinguished. Usually it is easy to state that, for instance, a given lexeme as a component of a noun phrase is unilaterally dependent on its constitutive member and, because of that, belongs to the class of adjectives. It is also fairly easy to qualify a lexeme having case government as a preposition. On the other hand, it is difficult to state if a lexeme can be used as a separate utterance and, therefore, can be included into the class of non-syntagmatic units. But many more difficulties are connected with the division of lexemes on the grounds of the connective function criterion. In numerous grammatical classifications one assumes that conjunctions, prepositions and relative pronouns have this function, while particles and adverbs do not. However, it is impossible to state if many units, e.g. bowiem ‘because’, jednak ‘however, nevertheless’, przecież ‘after all’, tymczasem ‘whereas, meanwhile’, satisfy in practice the definitional conditions for the connective function and, therefore, whether they belong, for instance, to the class of conjunctions or to particles. So it would be useful to look for new criteria (instead of the one mentioned above) coming into the domain of linear syntax.

Before the results of the word order inquiries are presented, one should consider a more general problem: are detailed classifications containing many small classes of units necessary? There are at least two arguments to support the affirmative answer to the question. Firstly, all the differences between grammatical features of lexical units ought to be described in the dictionaries of natural languages and, therefore, it is necessary to divide the whole set of the units in accordance with the criteria detailed earlier. Secondly, one can verify a general classification by means of many detailed divisions. Therefore, it would be possible to improve previous descriptions of the grammatical system of language.


2. My research on linear syntax deals only with synsyntagmatic lexical units in modern Polish. Nevertheless, they make a very heterogeneous set, which generates a number of methodological difficulties.

[139]Synsyntagmatic lexemes, as opposed to autosyntagmatic ones, do not occur in syntactic positions implied by other classes of language expressions, that is to say, they are not syntactically required but they have syntactic requirements for other expressions.

In this paper we accept the definition of synsyntagmatic lexemes mentioned above, as was recently proposed by Jadwiga Wajszczuk (cf. Wajszczuk, 1997, p. 52). There is a relevant difference between her interpretation of these lexemes and the one widely known from the handbook of Polish morphology (cf. Laskowski, 1984, 1998). In Laskowski’s opinion synsyntagmatic lexemes primarily do not function as sentence components and they obligatorily combine with autosyntagmatic lexemes or with syntactically complex structures. The denotations of the term under discussion in Laskowski’s and Wajszczuk’s opinions are different. Wajszczuk eliminates prepositions and relative pronouns from the class of synsyntagmatic lexemes in contrast to Laskowski, who included both of them into this class. On the other hand, particles are treated by Wajszczuk as belonging to the synsyntagmatic lexemes while the same type of units, which were called by Laskowski in Polish „modalizatory“, are classified in his works as autosyntagmatic lexemes.


3. In the paper the following basic theoretical assumptions concerning word order are accepted.


3.1. Word order as the relation of external arrangement of language expressions is based on the elementary notion of sequentiality and as such it is a two-argument relation.


3.2. Linear position of the lexical unit (in relation to the sentence or sentence component with which it is making a syntactic connection) is relevant for the syntactic classification of the unit. Description of the grammatical features of units ought to accompany the analysis of their linear positions.


3.3.1. A given unit X has a stable linear position in relation to a given expression Y, i.e. a sentence or a sentence component if only one pragmatically neutral linear arrangement of X in relation to Y is permissible.


3.3.2. A given unit X has a changeable position in relation to a given expression Y if at least two pragmatically neutral linear arrangements of X in relation to Y are possible. A linear changeability is conditioned not only by formal and functional features of the classes of the units but, first of all, by individual semantic features of the particular lexical units. Because of that, it is difficult to verify the word order rules.


3.3.3. If a given unit X has a stable linear position in relation to a given expression Y then X very often adjoins Y but that is not the necessary condition for the linear stability of the unit X.


3.4. Polysegmental lexical units especially the ones which do not contain obligatorily adjoining segments ought to be the subject of further research.


3.5. There are numerous extrasyntactic consequences of the change of linear arrangements, especially semantic, pragmatic and textual ones, but they do not directly concern the classification of lexical units based on the syntactic criteria.


[140]3.6. As for the theoretical assumptions of the word order description two more comments are needed.


3.6.1. Firstly, there are the different degrees of difficulties in establishing whether the linear position of a given unit is stable or changeable. It is relatively easy to state that a given unit X has a stable linear position if it enters a syntactic connection with only one sentence component which represents the definite grammatical class of units. For example, adnumerative operators of the type niespełna ‘somewhat less than, not quite’, bez mała ‘almost, nearly’ enter a relation with numerals or their derivatives used immediately after the operators. Cf. e.g.


Maria szła na przystanek autobusowy niespełna dwadzieścia minut. ‘It took Mary (slightly) less than 20 minutes to reach the bus stop.’


Jan czytał tę książkę przez bez mała (jeden) tydzień. ‘John was reading this book for almost a week (one week).’

Adsentential operators of the type jednak ‘however, nevertheless’, przecież ‘after all’, tymczasem ‘whereas, meanwhile’, widocznie ‘apparently’ stand in opposition to the units mentioned above. They enter syntactic relations with the sentence as a whole and they can be used in different linear positions, in both medial and initial positions. Cf. e.g.


Piotr miał wykształcenie muzyczne, jednak nigdy nie pracował w swoim zawodzie. ‘Peter had a degree in music, however, he has never worked in his profession.’


Piotr miał wykształcenie muzyczne, nigdy jednak nie pracował w swoim zawodzie.


Piotr miał wykształcenie muzyczne, nigdy nie pracował jednak w swoim zawodzie.


Przecież może pan kontynuować swoją wypowiedź, obrady zostały wznowione. ‘You can continue your speech after all, the session was resumed.’


Może pan przecież kontynuować swoją wypowiedź, obrady zostały wznowione.


Może pan kontynuować przecież swoją wypowiedź, obrady zostały wznowione.

Although the class of adsentential operators was discovered over one hundred years ago (A. Krasnowolski (1897) called it conjunctional adverbs), nobody has described the features of this class as a whole yet.


3.6.2. Secondly, one can interpret a linear changeability in different ways. However, in order to describe syntactic features of lexical units one should restrict the subject of examination to pragmatically neutral linear arrangements. If, as a result of the change of linear position of a given unit X, only the arrangements pragmatically marked arise, then one should state that the unit X has a stable position. For example, the inversion of non-contextual particles of the type prawie ‘almost’, zaledwie ‘only’ is often permissible, but it entails as a rule a change of pragmatic features of the sentence. Cf. e.g.


Bliźniaczki miały prawie identyczne fryzury. ‘The twin sisters had almost identical hairdos.’


Bliźniaczki miały identyczne prawie fryzury.


Ona wypiła zaledwie dwa kieliszki wina. ‘She drank only two wine-glasses.’


Ona wypiła dwa kieliszki wina zaledwie.

Therefore, this type of the word order change is not taken into further consideration.


[141]4. The central part of this paper is the following preliminary list of the subclasses of synsyntagmatic lexemes which were distinguished among others in accordance with word order criterion.


4.1. lexemes implying two syntactic positions

4.1.1. lexemes having a stable linear position: conjunctions occurring between two expressions joined by them (i.e. in the so-called interposition): conjunctions proper; e.g. i ‘and’, albo ‘or’, ale ‘but’, czyli ‘that is to say, in other words’ occurring before one of two expressions joined by them (i.e. in the so-called anteposition): auxiliary conjunctions; e.g. chociaż ‘although’, jeśli ‘if’, skoro ‘since’

4.1.2. lexemes having a changeable linear position in relation to the sequences occurring in the syntactic positions implied by them (by these lexemes): adsentential operators; e.g. bowiem ‘because’, jednak ‘however, nevertheless’, natomiast ‘whereas, while’,przecież ‘after all’, tymczasem ‘whereas, meanwhile’, widocznie ‘apparently’, wszak ‘after all’; some operators can only be used within the right-hand sequence (e.g. bowiem, jednak, natomiast, tymczasem), others within the right-hand as well as the left-hand sequence (e.g. przecież, widocznie, wszak) in relation to the sentence component with which they enter into syntactic connection: contextual particles; e.g. również ‘also, too’, także ‘also, too’, też ‘also, too’


4.2. lexemes implying one syntactic position

4.2.1. lexemes having a stable linear position before the sequence occurring in the syntactic position implied by them lexemes governing particular forms of verbs: mood operators; e.g. bodaj ‘I wish’, byle ‘I hope’, niech ‘let’, oby ‘may’ lexemes not governing particular verbal forms: quasi-conjunctional operators; e.g. alboż, ależ ‘why’, czyżby ‘oh?, really?’ before the sentence component with which they enter into a syntactic connection and which represents the definite grammatical class of units namely a numeral (and its derivatives): adnumerative operators, e.g. niespełna ‘somewhat less than, not quite’, około ‘about’, ponad ‘above, more than’, przeszło ‘above, more than’ a preposition: adprepositional operators; e.g. tuż ‘hard by, just after/before’, wprost ‘simply’, zaraz ‘at once, immediately’ and which does not represent one grammatical class of units lexemes entering a syntactic connection with a noun: non-contextual particles, e.g. chociaż ‘at least’, prawie ‘almost, nearly’, przynajmniej ‘at least’, zaledwie ‘only’ lexemes entering a syntactic connection with an adjective and a verb: [142]gradual operators, e.g. bardzo ‘very’, całkiem ‘quite’, dość ‘enough’, ledwie ‘only just, hardly’, nieco ‘a little, somewhat’, za ‘too’, zbyt ‘too’

4.2.2. lexemes having a changeable linear position in relation to the sentence component or the sentence with which they enter into a syntactic connection: modal operators; e.g. chyba ‘surely’, niejako ‘to some extent, in some measure’, pewnie ‘surely’, poniekąd ‘to some extent, in some measure’, właściwie ‘as a matter of fact’.


5.1. As for the list mentioned above, several comments and elucidations are needed. The division of synsyntagmatic units into (a) the ones implying two syntactic positions and (b) the others, implying only one syntactic position, is treated as superior in relation to the other features of the units. The second distinction of the units, namely, (c) the ones having a stable linear position and (d) the others, having a changeable one, is subordinated to the first subdivision and refers to both classes distinguished formerly. The remaining divisions are based on different criteria. The list of the subclasses of units proposed above is open, that is to say, one can distinguish new classes provided that new criteria of a division connected with a word order will be proposed.


5.2. It would be useful to compare the subclasses of units mentioned above with the widespread syntactic classifications in order to verify the hypotheses about word order and to formulate new ones.

Conjunctions, particles and adverbs constitute internally heterogeneous sets from the viewpoint of linear arrangement.

5.2.1. Only conjunctions proper and so-called auxiliary conjunctions, that is, conjunctions in the strict sense, have stable linear positions. Both groups of the units were recently distinguished, described and opposed to so-called quasi-conjunctions by Jadwiga Wajszczuk (Wajszczuk, 1997). The last-mentioned, qualified in Polish linguistic literature as conjunctions or particles, are treated in this paper as a separate subclass of units called adsentential operators. They have changeable linear positions in relation to the sequences which occur in syntactic positions implied by these operators. Cf. e.g. (3)–(8).

5.2.2. Contrary to widespread opinions (cf. Laskowski, 1984, 1998; Wróbel, 1996; Grochowski, 1986, 1997; Maldjieva, 1995) positional variability of particles is not a typical feature for all of the units belonging to this class. One should divide them at least into three subclasses: (a) contextual particles, (b) non-contextual particles and (c) modal operators. The units of the first subclass, in contrast to both remaining, imply two syntactic positions and in this respect they are similar to conjunctions and adsentential operators. As opposed to the latter (b and c), however, they enter syntactic connections (like non-contextual particles) with the sentence component and they occur in different linear positions in relation to it. Cf. e.g.


Również Joanna jest mężatką (a nie tylko moja starsza córka). ‘Also Jane is a married woman (and not only my elder daughter).’


Joanna również jest mężatką (a nie tylko moja starsza córka).

Non-contextual particles have a stable linear position, namely, they precede the sentence component with which they enter into a syntactic connection. Cf. e.g.


Jan ma włosy niemal do ramion. ‘John’s hair almost reaches his arms.’


Maria przegląda w gazetach przynajmniej nekrologi. ‘Mary looks through at least the obituary column in the newspapers.’

Positional variability of modal operators is unquestionable, but it is difficult to state without hesitation whether they enter syntactic connections with sentence components or with the sentence as a whole. Cf. e.g.


Pewnie jej rodzice sprzedali samochód. ‘Surely her parents sold the car.’


Jej rodzice pewnie sprzedali samochód.

5.2.3. A great many adverbs do not belong to synsyntagmatic lexemes, but only to autosyntagmatic ones, and because of that, adverbs in principle were excluded from the field of my research. However, there are adverbs which fulfil the definitional conditions for synsyntagmatic lexemes, among others, they are so-called gradual operators. They are used as antecedents of adjectives, verbs and adverbs with which they enter syntactic connections. Cf. So there is an incompatibility between the last-mentioned feature of gradual operators and the positional variability of typical adverbs.

5.2.4. In addition, the list of synsyntagmatic lexemes mentioned above contains four subclasses of units having stable linear positions, i.e. (a) mood operators, (b) quasi-conjunctional operators, (c) adnumerative operators, (d) adprepositional operators. In traditional classifications the first three sets of operators were included into the class of particles, the fourth one into the class of adverbs. The units belonging to the first two sets of operators occur always in the initial position. Adnumerative operators are used immediately before a numeral or its derivative while adprepositional ones immediately before a preposition. All of the four subclasses of operators have been distinguished and described in my earlier works (cf. Grochowski, 1986, 1997).


6. The considerations presented in this paper lead to two most general conclusions.


6.1. All synsyntagmatic lexemes having a stable linear position precede the expression (the sentence or the sentence component) with which they enter into a syntactic connection. Only conjunctions proper are exceptions to this general rule. They are placed between two expressions joined by them.


6.2. Conjunctions, particles and adverbs are heterogeneous sets from the viewpoint of linear arrangement of units belonging to them. In this connection, further subdivisions of the sets mentioned above are needed. Especially adsentential operators, contextual particles and modal operators ought to be further examined.




GROCHOWSKI, M.: Polskie partykuły. Składnia, semantyka, leksykografia. Wrocław 1986.

GROCHOWSKI, M.: Wyrażenia funkcyjne. Studium leksykograficzne, Kraków 1997.

KRASNOWOLSKI, A.: Systematyczna składnia języka polskiego. Warszawa 1897.

LASKOWSKI, R.: Podstawowe pojęcia morfologii. In: Gramatyka współczesnego języka polskiego. Morfologia. Ed. R. Grzegorczykowa – R. Laskowski – H. Wróbel. Warszawa 1984, p. 9–57.

[144]LASKOWSKI, R.: Zagadnienia ogólne morfologii. In: Gramatyka współczesnego języka polskiego. Morfologia. Warszawa 1998, p. 27–86, 2nd ed.

MALDJIEVA, V.: Non-inflected Parts of Speech in the Slavonic Languages. Syntactic Characteristics. Warszawa 1995.

SALONI, Z.: Klasyfikacja gramatyczna leksemów polskich. Język Polski, 54, 1–2, p. 3–13, 93–101.

WAJSZCZUK, J.: System znaczeń w obszarze spójników polskich. Wprowadzenie do opisu. Warszawa 1997.

WRÓBEL, H.: Nowa propozycja klasyfikacji syntaktycznej polskich leksemów. In: Studia z leksykologii i gramatyki języków słowiańskich. Ed. H. Wróbel. Kraków 1996, p. 53–60.



Ke vztahu pořádku slov a synsyntagmatických lexémů v polštině

Stať klade otázku, do jaké míry lze použít pořádek slov jako kritérium syntaktické klasifikace lexikálních jednotek. Zkoumání je zaměřeno na synsyntagmatické lexikální jednotky polské slovní zásoby. Autor dospívá k závěru, že:

1. Všechny synsyntagmatické lexémy mající stabilní lineární pozici předcházejí výraz (tj. větu nebo větný komponent), se kterým vstupují do syntaktických vztahů. Výjimkou jsou pouze vlastní spojky (conjunctions proper). Jsou umístěny mezi dvěma výrazy, které spojují.

2. Spojky, částice a adverbia jsou z hlediska lineárního uspořádání heterogenní množinou. V této souvislosti je nezbytná jejich další subklasifikace. V budoucnosti musejí být zkoumány zvláště větné operátory (adsentential operators), kontextové částice (contextual particles) a modální operátory.

Instytut Języka Polskiego, Universytet Mikolaja Kopernika
Fosa Staromiejska 3, Toruń, Polska

Slovo a slovesnost, ročník 61 (2000), číslo 2, s. 138-144

Předchozí Jasňa Šlédrová: K sémantice českých rozměrových adjektiv

Následující Jiří Kraus: Pozoruhodný slavistický příspěvek k pragmatice spojovacích výrazů