Processing and Representation of Different Types of Czech Affixes

Processing and Representation of Different Types of Czech Affixes Radka Julínková — Denisa Bordag ABSTRACT: The study investigates the processing of morphologically complex words in Czech. In Experiment 1 we employed morphological repetition priming to test the Split Morphology Hypothesis, i.e. whether derived and inflected word forms are stored in the same or different manner in the Czech mental lexicon. The results demonstrate significantly larger priming effects for inflected forms compared to derived forms indicating distinct processing of inflection and derivation in Czech; while inflected forms are fully decomposed during language comprehension, derived forms are either not, or only partially. In Experiment 2 we addressed two research questions. First, we tested the psycholinguistic reality of the linguistic distinction between two types of inflective verbal prefixes: (a) “purely” inflective aspectual prefixes (i.e. the prefix turns an imperfective verb into a perfective one as in hřešit (imp.; ‘to sin’) — zhřešit (perf.)) and (b) derivational verbal prefixes (e.g. krátit (imp.; ‘to shorten’) — zkrátit (perf.)). The results did not indicate any evidence that this distinction would be psycholinguistically grounded. Second, we examined the role of semantic transparency of the derivational prefixes in the processing. The experiment delivered evidence of slower processing of opaque derived verbs, most likely caused by double search/reanalysis. KEY WORDS: affixes, aspect, Czech, derivation, inflection, lexical decision, priming, semantic transparency INTRODUCTION The processing of morphologically complex words and the nature of their mental representation have been hotly discussed topics in psycholinguistic debates during the last decades. Is a morphologically complex word decomposed before being accessed? Are affixes represented independently of the stem? Does the processing of different types of affixes proceed in a different manner? These and similar questions have been addressed in several languages, however without conclusive results. Some studies suggest that complex words are accessed in their full forms, especially if they are concerned with semantically opaque (Sandra, 1990; Monsel, 1985; Marslen-Wilson et al., 1994), morphologically irregular (Ullman, 2001; 2004; Ullman et al., 2002; Pinker, 1997; 1998; 1999; Pinker — Ullman, 2002; Clahsen — Aveledo — Roca, 2002; Münte et al., 1999; Newman et al., 2007; Allen — Badecker — Osterhout, 2003), or very frequent and common inflected forms (Monsel, 1985; Sandra, 1990). On the contrary, other findings support morphological decomposition, i.e. accessing words through their components, particularly in the case of transparent compounds (Jarema et al., 1999; Zwitserlood, 1994; Sandra, 1990) and/or transparent complex words (Marslen-Wilson et al., 1994) or regular inflected RADKA JULíNKOVá — DENISA BORDAG 53 forms (Ullman et al., 2002; Ullman, 2004; Pinker, 1999; Münte et al., 1999; Newman et al., 2007; Allen — Badecker — Osterhout, 2003). In general, there is no agreement about how morphologically complex words are accessed, both across languages and with respect to various phenomena within one language (for review of conflicting results see McQueen — Cutler, 1998). What previous research did show with certainty is that results from one language cannot be readily-generalized to other languages. For example, the so-called Split Morphology Hypothesis (Anderson, 1988) claiming that inflection and derivation are two different mental processes, was found valid in English (Stanners et al., 1979), Dutch (Bertram et al., 2000), Finnish (Bertram — Laine — Karvinen, 1999) and German (Schriefers — Friederici — Graetz, 1992), however, could not be confirmed for Hebrew (Feldman — Bentin, 1994). A linguistic phenomenon that is cross-linguistically identified as “the same” can thus be underlied by different morphological processes and mental structures, making the testing of a broad scope of individual languages a sine qua non for verifying general claims. Independently from this fact, different languages open different possibilities for testing both new and old hypotheses about language processing and representation (e.g. Bordag — Pechmann, 2009). The psycholinguistic examination of the Czech language as presented in this study can thus provide unique insights both into morphological processing in general and into a specific language with a rich inflectional system with various types of affixes in particular. In our study we focus on the processing of selected types of affixes and their representation in the Czech mental lexicon. PROCESSING OF MORPHOLOGICALLY COMPLEX WORDS As indicated in the introduction, there are several hypotheses concerning processing morphologically complex words. According to the so-called Affix Stripping Hypothesis introduced by Taft and Forster (1976), all affixes are stripped at first and then the stem decoding takes place (Rastle — Davis — New, 2004). This Obligatory Decomposition Hypothesis (also e.g. Smith — Sterling, 1982; Taft, 2004) is supported by the results of Taft’s (1979; 1981) experiments revealing that the processing of words with the initial part resembling prefixes (e.g. revise) takes longer than processing of non-prefixed words (e.g. divide). According to Taft, the finding indicates that affix stripping takes place even in such cases: The presumed affix reis first stripped, but since there is no entry for vise, the word form has to be reanalysed, which delays the processing. The Affix Stripping Hypothesis assumes links of various strength and character among morphemes constituting morphologically complex words and predicts processing differences with respect to the type of affix (Kiparsky, 1982). A modification of this hypothesis by Colé, Segui and Taft (1997) proposed that the search for complex and decomposed forms proceeds in parallel and that the frequency of the complex form, on the one hand, and of its stem, on the other, crucially determines which search will be faster. The finding that access to complex word forms with high frequent stems is faster than to complex words forms with low frequent stems also supports the 54 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2015 Decomposition Hypothesis (Taft, 1979; 2004; in compounds: Fiorentino — Poeppel, 2007). Further evidence for the decomposition of morphologically complex words comes from studies using masked priming (Rastle — Davis — New, 2004; Lehtonen et al., 2011) or combining masked priming and magnetoencephalography (MEG) (e.g.Solomyak — Marantz, 2010). The so-called Full Listing Hypothesis (Butterworth, 1983) represents the opposite view assuming that complex words are represented in their full forms without any reference to their constituents. This view was seen as uneconomical with respect to the used capacity in the mental lexicon by listing all inflectional words in the declarative memory. However, decomposition and construction of complex forms puts functional load on working memory in turn. Moreover, recent approaches suggest that significantly more is stored in the memory than originally assumed, including not only frequent full forms, but also frequent chunks of words (Langacker, 1987; Goldberg, 2003; Hanna — Pulvermüller, 2014). Dual Pathway Models (e.g. Marslen-Wilson et al., 1994; Wurm, 1997) is a term often used for approaches combining both views, i.e. morphological decomposition for some words (typically regular and transparent) and non-decomposition for others. Whether complex words are decomposed or retrieved from memory as complex units is thus related to the nature of their constituents. Also, supporters of morphological decomposition typically agree that some words, such as irregular verbs (Pinker, 1999; Ullman et al., 2002; Ullman, 2004; Münte et al., 1999; Newman et al., 2007; Allen — Badecker — Osterhout, 2003; Clahsen — Sonnenstuhl — Blevins, 2003; Sonnenstuhl — Eisenbeiss — Clahsen, 1999), forms involving alternations (Caramazza et al., 1985) or even some frequent, common inflected forms and opaque compounds might be stored in the form of a whole nondecomposed word (Monsel, 1985; Sandra, 1990; Bertram — Schreuder — Baayen, 2000; Baayen et al., 2003). The experiments of Ito, Sugioka and Hagiwara (1996) and Hagiwara et al. (1999) with Japanese complex words indicate that the decomposition process takes place only in regularly affixed words because only then can a rule be applied (cf. also Beretta et al., 2003; an fMRI study with German). Their claims are based on the comparison of data obtained from Broca and Wernicke’s aphasics; while Broca’s aphasics showed difficulties in the production of regular derivational suffixes, but not in production of irregular ones, the pattern was reversed for Wernicke’s aphasics. Vannest and Boland (1999) suggested (based on results for English) a division of affixes into two groups: one comprising productive, phonologically neutral, semantically transparent affixes yielding the decomposition process (e.g. -less; worthless) and the other with idiosyncratic, structure changing, semantically-opaque affixes that are not decomposed (e.g. -ity; severity). Related to the above considerations is the Split Morphology Hypothesis (Anderson, 1977; 1982; 1988; 1992; Matthews, 1972; 1991; Perlmutter, 1988; Scalise, 1984; 1988), according to which derivation and inflection are two different mental processes because two different types of constituents/affixes are involved in these operations. Derivation is considered a word formative process in which a newly-formed word is conceived as a different lexical item with a different meaning (at least to a certain RADKA JULíNKOVá — DENISA BORDAG 55 extent). The derived word also typically belongs to a different word class. On the other hand, inflected forms are considered forms of the same lexeme, i.e. they are variations of the same word and thus have the same lexical meaning. Inflection is predictable because it is (with some exceptions) regular and rule-governed, which cannot be clearly claimed in the case of derivation. Inflectional forms are also more constrained by syntax than derivational forms (Anderson, 1982). PRESENT STUDY In the present study two experiments are introduced. The aim of Experiment 1 was to test the potential processing differences between inflected and derivational forms in relation to the Split Morphology Hypothesis in Czech. In Experiment 2, we explored the processing of prefixed verbs with presumably different statuses of their prefixes with respect to inflection and derivation. We intended to test whether the particular linguistic analysis of verbal prefixes corresponds to their mental representation or whether it is rather a linguistic construct not corresponding to any processing correlate. In addition, we also investigated the role of semantic transparency/opacity in the processing of various derivational prefixes. SPLIT MORPHOLOGY HYPOTHESIS (SMH) IN RELATION TO INFLECTION AND DERIVATION Our first research question (Experiment 1) was to test SMH in Czech with the main purpose of finding out whether there is a difference between the processing of inflection and derivation in this morphologically very rich language. SMH has been tested, for instance, in English, Italian and German with results indicating different processing of inflection and derivation (Stanners et al., 1979; Fowler — Napps — Feldman, 1985; Laudanna — Badecker — Caramazza, 1992; Schriefers — Friederici — Graetz, 1992). However, as mentioned earlier, the results from Hebrew (Feldman — Bentin, 1994) did not provide any evidence of derivation and inflection as two mentally different processes, which might be ascribed to the non-concatenative nature of Hebrew morphology. As the nature of a language’s morphology obviously plays a role, the hypothesis should not be generalized over different languages before being tested. Slavonic languages, in general and Czech in particular, possess a rich inflectional system with frequent stem changes due to alternations. To our knowledge, the only experiment testing SMH in Slavonic languages was conducted in Serbian by Feldman (1994). Her results provide evidence that the processing of inflection and derivation differs in Serbian, and thus confirm the validity of SMH in this Slavonic language. However, as previous research shows (e.g. Bordag — Pechmann, 2009), evidence from one language of a particular language family does not ensure validity for all languages in the given family and replications are needed to assess the generalizability and the scope of the hypothesis. 56 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2015 Our experiment is thus a partial replication of Feldman (1994) with several adjustments in design. Like Feldman, we examine the processing of inflected and derived nouns and verbs and focus on the facilitation in processing of morphologically related forms. EXPERIMENT 1 The experiment is an adjusted replication of Feldman (1994). Using morphological repetition priming (identical, inflectional, derivational) with lexical decision, we tap into processing of the derivational and inflectional forms to find out whether different types of suffixes are processed differently and whether SMH (claiming that derivation and inflection are different mental processes) is valid also for the Czech language. PARTICIPANTS Twenty-three Czech native speakers, all right-handed university students (9 men, 14 women, aged 20–30 years), participated in the experiment. They were remunerated for participating in the study. STIMULI A battery suitable for Czech speakers was constructed from nouns and verbs in three forms (basic form, inflectionally related form, derivationally related form).1 Twenty-seven Czech word triples were selected. Fifteen triples consisted of a noun target in the nominative (basic form that served also as an identity prime), an inflectionally related form in instrumental and a derivationally related verb form. That is for instance: the nominative target HMAT (‘a touch’) constitutes a pair with prime in the instrumental hmatem and with the derivationally-related verb hmatáš (‘you touch’) in the 2nd person singular indicative present. Twelve triples consisted of a verb target in the 1st person singular indicative present (basic form) and an inflectionally related form in the 2nd person singular indicative present and a derivationally-related agentive the noun form. For example, target PLAVU (‘I swim’) made a triple with plaveš (‘you swim’) and plavec (‘a swimmer’). Twenty-seven triples of orthographically and phonemically regular pseudowords were created by changing one or two letters in real words (vowel for vowel, consonant for consonant). Triplets were organized according to the same principle as real words, i.e. pseudo-words were declined and conjugated as if they were real words, i.e. contained suffixes of existing words. A further 162 pairs of fillers were constructed in order to prevent expectancy and strategic effects. The primes and targets were always semantically and 1 Only masculine nouns and verb forms of all verbal classes were included in the testing battery. RADKA JULíNKOVá — DENISA BORDAG 57 morphologically unrelated. They were divided into three subgroups of 54 items. In the first group the prime and the target were of the same type (i.e. word — word or pseudo-word — pseudo-word). For example, word — word pairs were such as DĚLÁM (‘I do’) — PATŘÍM (‘I belong to’) or HROZEN (‘a grape’) — SEVER (‘north’). The pair members of the items in the second group were of a different type with the primes being words and the targets pseudo-words, or vice versa. Thus TEPLO (‘warm’) — KODLOM and HUNKUJU — BRUSLÍM (‘I’m skating’) were matched. Items in the last group were paired in the same manner as in the second group, but there was a phonological overlap between their first syllables that were identical e.g. FARMÁŘ (‘farmer’) — FARPON, or TÉMA (‘theme’) — TÉNO. All items included in the battery were 3–7 letters long. Primes were 1–2letters longer than targets and both were perfectly matched phonemically and orthographically. To ensure perfect overlap, neither alternations nor softening or shortening were included. Ambiguity of any sort was avoided as well. All words were assumed highly familiar for Czech native speakers. PROCEDURE The morphological priming experiment was designed in E-Prime software (E-Prime, v. 1.2; Schneider — Eschman — Zuccolotto, 2002). Each test consisted of 684 prime-target pairs, of which 162 items were of interest. At first, a prime appeared on the screen for 200 ms, followed by a fixation sign for 200 ms and finally by a target for 800 ms. Primes appeared in the middle of the screen in lower-case letters and targets in capital letters (to avoid graphical overlap and highlight the items to which participants should respond), both printed in Courier New script 18. Participants performed lexical decision on the target by determining whether it is an existing Czech word by pressing the YES or NO button on a keyboard. Reaction times on targets were measured. Primes were either identical with targets or they were their inflections or derivations. This approach, rather than presenting the inflected/derived forms as targets, was taken to avoid confusion about the lexical status of the inflected forms (especially for nouns) presented without context. The experimental paradigm can be illustrated as following: The battery was divided into three blocks which were randomized into three different orders. The initial six items in each block served as fillers and were not included in the analysis. Items in each block were randomized automatically. One testing lasted approximately 25 minutes. Before the test phase participants performed a brief practice phase consisting of 12 items. 58 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2015 RESULTS AND DISCUSSION Reaction times slower than 680 ms (17.9%) were excluded from the analysis as outliers on the basis of Feldman’s (1994) experiment. Reaction times for identity, inflectional and derivational priming with words are presented in Table 1. One factorial analyses of variance (ANOVAs) with the factor Type of Prime (three levels: identity, inflectional, derivational) was performed over the target latencies for the critical items using subjects (F1) and items (F2) as random variables. The effect of Type of Prime revealed a difference in the reaction times between the three groups: F1(46,2) = 4.66, p <.05; F2(52,2) = 5.78, p <.01. A Scheffé test revealed that the reaction times in the identity and inflectional condition were statistically the same, but that they both significantly differed from the derivational condition (p <.05) where the reaction times on the target were slower. This suggests different processing of inflectional and derivational forms. In other words, the priming effect due to derivationally-related primes was significantly weaker than the priming effect due to inflectionally related primes. Primes Example Identity prime Inflectional prime Derivational prime PLAVU (1.Sg.Pres.Ind.) PLAVU (1.Sg.Pres.Ind.) PLAVU (1.Sg.Pres.Ind.) PLAVU (1.Sg.Pres.Ind.) PLAVEŠ (2.Sg.Pres.Ind.) PLAVEC (agentive noun) or HMAT (nominativ) HMAT (nominativ) HMAT (nominativ) HMAT (nominativ) HMATEM (instrumental) HMATÁŠ (related verb) Word RT (ms) Nr. ERR (%) RT (ms) Nr. ERR (%) RT (ms) Nr. ERR (%) 524.86 530 3 524.36 514 4 536.45 521 4 table 1: RTs and error rates with real-word targets. The statistically identical identity and inflectional priming might be interpreted as a result of accessing the same lexical entry for all inflected word forms. Smaller priming effect observed in derivationally primed items suggests access involving a more complex structure that e.g. employs more nodes or different types of links among constituents as derived word forms might be only partially decomposed or might not be decomposed at all. In the latter case, two different entries would be accessed in the mental lexicon. Analyses including word class (nouns vs. verbs) as a factor revealed that nouns were processed significantly faster than verbs (F1(1,22) = 11.9, p <.01; F2(1,24) = 8.6, p<.01 and that there was no difference in the priming between the two word classes (Fs < 1) (see Table 2). Faster processing of nouns in words might result from the fact that target nouns were in basic forms in the nominative singular, while target verbs were in the 1st person singular present. Thus, while noun target had a zero final morpheme, verb targets involved an inflectional suffix. This fact might be considered only in the processing of identity and inflectional forms. RADKA JULíNKOVá — DENISA BORDAG 59 Word Pseudoword Target Noun Verb Noun Verb Identity priming 514.94 537.22 567.05 582.77 Inflectional priming 517.40 532.95 566.48 588.61 Derivational priming 531.15 543.42 569.60 585.26 table 2: Mean RTs for nouns and verbs in real words and pseudo-words conditions. Separate analyses were performed over the control filler pseudo-words revealing no significant differences between the three conditions (see Table 3). This finding shows that the observed differences between inflection and derivation are of a morphological nature and not an artefact of the employed suffixes: When attached to pseudo-words, no difference in priming effect was observed showing that the orthographic overlaps alone cannot account for the results with existing words. Example Primes Identity prime Inflectional prime Derivational prime PIRU (1.Sg.Pres.Ind.) PIRU (1.Sg.Pres.Ind.) PIRU (1.Sg.Pres.Ind.) PIRU (1.Sg.Pres.Ind.) PIREŠ (2.Sg.Pres.Ind.) PIRAN (agentive noun) or KEZ (nominativ) KEZ (nominativ) KEZ (nominativ) KEZ (nominativ) KEZEM (instrumental) KEZÍŠ (related verb) Pseudoword RT (ms) Nr. ERR (%) RT (ms) Nr. ERR (%) RT (ms) Nr. ERR (%) 573.53 342 11 574.96 329 13 575.40 359 10 table 3: Mean RTs and error rates with pseudo-word targets. The experiment confirmed the results of previous studies in English, German, Dutch, Italian and Serbian: priming was significantly smaller when the prime was a derived, rather than an inflected, form. The finding indicates that derivation and inflection are two distinct mental processes in Czech as well. THE MORPHOLOGICAL STATUS OF CZECH VERBAL PREFIXES The status of verbal prefixes with respect to derivation and inflection is highly controversial in Czech. Most research agrees that verbal prefixes typically combine two functions: (a) lexical (they are considered to be word-forming means), and (b) grammatical (they mark aspectual changes). The controversy concerns especially a group of verbs whose prefixes are purely grammatically or lexically empty according to some research (e.g. Kopečný, 1962; Šlosar, 1981; 1986). According to these authors, the important property of verb forms with such prefixes is their inability 60 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2015 to be turned into secondary imperfective forms (or the frequency of these forms is very low). As an example, the verb ničit (‘to keep destroying’) can be turned into a perfective verb by adding the prefix z-, i.e. zničit (‘to destroy’), but the secondary imperfective form *zničovat is ungrammatical. On the other hand, perfective verbs with prefixes that also have a lexical function can be turned into secondary imperfective verbs, e.g. zkrátit (‘to shorten’), utlumit (‘to deaden’) and these again into secondary imperfectives zkracovat, utlumovat. Thus, although prefixation is generally considered to be a word formation process (new words are produced, not just new forms of a word), in purely aspectual prefixation prefixes do not form new lexemes, but only change the grammatical function of the given verb. According to this approach, the lexical prefixes are classified as derivation, while the purely aspectual affixes are classified as inflection. The controversial status of verbal prefixes, in general, and the existence of purely aspectual prefixes, in particular, has been a subject of discussion for many years not only in Czech, but in other languages such as Russian. Endresen et al. (2012) argue that the traditional assumption that prefixes are semantically “empty” when used to form aspectual pairs is problematic because the same prefixes are clearly “non-empty” when combined with other base verbs (in Czech the example of a semantically empty prefix can be found in skončit (‘to end’), however, the prefix scannot be declared as empty in smést (‘to sweep away’, ‘to sweep together’) as it expresses the meaning ‘away from the surface’ and ‘together’). Endresen et al. (2012) also proposed that prefixes are not empty, but instead have meanings that overlap with the meanings of the base verbs. This is valid also for Czech. Uher (1987) claimed (about Czech) that verbal prefixes do not express perfective aspect only even in “purely” aspectual forms and therefore it is the most adequate to consider aspect as a lexically-grammatical category (“Slovesný vid je proto nejvhodnější považovat za lexikálně-gramatickou kategorii […]” (Uher, 1987, p. 46).) Other linguists (e.g. Maslov, 1963; Komárek, 2006; Veselý, 2010) also asssume that “purely” aspectual prefixes do not exist. A different opinion is held, for instance, by Kopečný (1962) and Šlosar (1981; 1986). According to their views, some purely aspectual prefixes do not express any lexical meaning and some express a semantic feature which already constitutes a part of semantics of the verbal meaning itself. In these cases the prefix appears lexically empty and “purely” aspectual (“Prefix vyjadřuje významový rys, který je součástí sémantiky slovesného významu samého. V takových případech se prefix jeví navenek jako lexikálně prázdný, prostě vidový” (Šlosar, 1986, p. 339).) Veselý (2010), who sees a contradiction in the statement itself, disagreed with this claim. According to him, it is not possible to claim simultaneously that a certain prefix has and, at the same time, does not have a lexical meaning (because the meaning is repetitive or implicitly present in a fundamental verb) (“Není možné současně tvrdit, že určitá předpona lexikální význam má i nemá. Nelze tvrdit, že ho vlastně jakoby nemá proto, že jde o význam opakovaný (nebo implicitně přítomný i ve fundujícím slovese)” (Veselý, 2010, p. 118).) In Experiment 2, we explored whether the distinction between derivational and purely aspectual affixes proposed by some authors is psycholinguistically grounded. RADKA JULíNKOVá — DENISA BORDAG 67 verbs are processed faster because both the direct and the decomposition route contribute to the word recognition. The retrieval of opaque prefixed verbs is delayed due to the conflicting outcome of the direct route and the decomposition route, which misleadingly attempts to construct the meaning of the complex word from the meanings of its individual constituents. Schreuder and Baayen’s model also includes a mechanism of activation feedback between the lemma nodes (linked with a syntactic and semantic layer) and the representations of constituents that are fully present in the complex words. The mechanism allows cumulative frequency effects for transparent complex words, but not for opaque ones. Gradually, the activation feedback tunes the system towards an advantage for the decomposition route which also results in a processing benefit for the transparent, but not the opaque words. Our study represents the first attempt to address the differences in processing of derivational and inflectional suffixes and prefixes in Czech, as well as the role of transparency/opacity in the processing of complex words. Further research is necessary to extend our knowledge about the processing of complex language forms, both in Czech and in other (often rarely explored) languages. ACKNOWLEDGEMENTS This research was supported by the student project of Palacký University (IGA: “Verbální prefixy a jejich role při ukládání a organizaci položek v mentálním lexikonu”; SPP 432101601; Experiment 2) and by “Lingvistická a lexikostatistická analýza ve spolupráci lingvistiky, matematiky, biologie a psychologie” (CZ.1.07/2.3.00/20.0161), co-financed by European social fund and state budget of the Czech Republic (Experiment 1). REFERENCES: Allen, Mark — Badecker, William — Osterhout, Lee (2003): Morphological analysis during sentence processing: an ERP study. Language and Cognitive Processes, 18(4), pp. 405–430. Anderson, Stephen R. (1977): On the formal description of inflection. Proceedings of Chicago Linguistic Society, 13, pp. 15–44. Anderson, Stephen R. (1982): Where’s morphology? Linguistic Inquiry, 13(4), pp. 571–612. Anderson, Stephen R. (1988): Morphological theory. In: Frederick J. Newmeyer (ed.), Linguistics: The Cambridge Survey: 1: Linguistic Theory: Foundations. Cambridge: Cambridge University Press, pp. 146–191. Anderson, Stephen R. (1992): A-Morphous Morphology. Cambridge: Cambridge University Press. Baayen, R. Harald — McQueen, James — Dijkstra, Ton — Schreuder, Robert (2003): Frequency effects in regular inflectional morphology: revisiting Dutch plurals. In: R.Harald Baayen — Robert Schreuder (eds.), Morphological Structure in Language Processing. Berlin: Mouton, pp. 355–390. Beretta, Alan — Campbell, Carrie — Carr, Thomas H. — Huang, Jie — Schmitt, Lothar M. — Christianson, Kiel — Cao, Yue (2003): An ER-fMRI investigation of morphological inflection in German reveals that the brain makes a distinction between 68 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2015 regular and irregular forms. Brain and Language, 85(1), pp. 67–92. Bertram, Raymond — Laine, Matti J. — Baayen, R. Harlad — Schreuder, Robert — Hyönä, Jukka (2000): Affixal homonymy triggers full form storage, even with inflected words, even in a morphologically rich language. Cognition, 74(2), pp. B13–B25. Bertram, Raymond — Laine, Matti J. — Karvinen, Katja (1999): The interplay of word formation type, affixal homonymy, and productivity in lexical processing: evidence from a morphologically rich language. Journal of Psycholinguistic Research, 28(3), pp. 213–226. Bertram, Raymond — Schreuder, Robert — Baayen, R. Harald (2000): The balance storage and computation in morphological processing: the role of word formation type, affixal homonymy, and productivity. Journal of Experimental Psychology: Learning, Memory, and Cognition, 26(2), pp. 489–511. Bordag, Denisa — Pechmann, Thomas (2009): Externality, internality, and (in-)dispensability of grammatical features in speech production: evidence from Czech declension and conjugation. Journal of Experimental Psychology: Learning, Memory, and Cognition, 35(2), pp. 446–465. Butterworth, Brian (1983): Lexical representation. In: Brian Butterworth (ed.), Language Production: Volume 2: Development, Writing and Other Language Processes. London: Academic Press. Caramazza, Alfonso — Miceli, Gabriele — Silveri, M. Caterina — Laudanna, Alessandro (1985): Reading mechanisms and the organization of the lexicon: evidence from acquired dyslexia. Cognitive Neuropsychology, 2(1), pp. 81–114. Clahsen, Harald — Aveledo, Fraibet — Roca, Iggy (2002): The development of regular and irregular verb inflection in Spanish child language. Journal of Child Language, 29(3), pp. 591–622. Clahsen, Harald — Sonnenstuhl, Ingrid — Blevins, James P. (2003): Derivational morphology in the German mental lexicon: a dual mechanism account. In: R. Harald Baayen — Robert Schreuder (eds.), Morphological Structure in Language Processing [Trends in Linguistics, Studies and Monographs, 151]. Berlin — New York, NY: Mouton de Gruyter, pp. 125–155. Colé, Pascale — Segui, Juan — Taft, Marcus (1997): Words and morphemes as units for lexical access. Journal of Memory and Language, 37(3), pp. 312–330. Endresen, Anna — Janda, Laura A. — Kuznetsova, Julia — Lyashevskaya, Olga — Makarova, Anastasia — Nesset, Tore — Sokolova, Svetlana (2012): Russian ‘purely aspectual’ prefixes: not so ‘empty’ after all? Scando-Slavica, 58(2), pp. 231–291. Feldman, Laurie B. (1994): Beyond orthography and phonology: differences between inflections and derivations. Journal of Memory and Language, 33(4), pp. 442–470. Feldman, Laurie B. — Bentin, Shlomo (1994): Morphological analysis of disrupted morphemes: evidence from Hebrew. Quarterly Journal of Experimental Psychology: Section A: Human Experimental Psychology, 47(2), pp. 407–435. Fiorentino, Robert — Poeppel, David (2007): Compound words and structure in the lexicon. Language and Cognitive Processes, 12(7), pp. 953–1000. Fowler, Carol A. — Napps, Shirley E. — Feldman, Laurie B. (1985): Relations among regular and irregular morphologically related words in the lexicon as revealed by repetition priming. Memory and Cognition, 13(3), pp. 241–255. Giraudo, Hélène — Grainger, Jonathan (2000): Effects of prime word frequency and cumulative root frequency in masked morphological priming. Language and Cognitive Processes, 15(4–5), pp. 421–444. Goldberg, Adele E. (2003): Constructions: a new theoretical approach to language. Trends in Cognitive Sciences, 7(5), pp. 219–224. Hagiwara, Hiroko — Sugioka, Yoko — Ito, Takane — Kawamura, Mitsuru — Shiota, Junichi (1999): Neurolinguistic evidence for rule-based nominal suffixation. Language, 75(4), pp. 739–763. RADKA JULíNKOVá — DENISA BORDAG 69 Hanna, Jeff — Pulvermüller, Friedemann (2014): Neurophysiological evidence for whole form retrieval of complex derived words: a mismatch negativity study. Frontiers in Human Neuroscience, 8, pp. 886–898. Ito, Takane — Sugioka, Yoko — Hagiwara, Hiroko (1996): Psychological status of rules in derivational morphology: evidence from Japanese nominal suffixation. Metropolitan Linguistics, 16, pp. 18–30. Jarema, Gonia — Busson, Céline — Nikolova, Rossitza — Tsapkini, Kyrana — Libben, Gary (1999): Processing compounds: a cross-linguistic study. Brain and Language, 68(1–2), pp. 362–369. Kiparsky, Paul (1982): Lexical phonology and morphology. In: In-Seok Yang (ed.), Linguistics in the Morning Calm. Seoul: Hanshin, pp. 3–91. Komárek, Miroslav (2006): Prefixace a slovesný vid (K prefixům prostě vidovým a subsumpci). In: Miroslav Komárek, Příspěvky k české morfologii. Olomouc: Periplum, pp. 183–195. Kopečný, František (1962): Slovesný vid v češtině. Praha: Československá akademie věd. Langacker, Ronald W. (1987): Foundations of Cognitive Grammar: Volume I: Theoretical Prerequisites. Stanford, CA: Stanford University Press. Laudanna, Alessandro — Badecker, William — Caramazza, Alfonso (1992): Processing inflectional and derivational morphology. Journal of Memory and Language, 31, pp. 333–348. Lehtonen, Minna — Monahan, Philip J. — Poeppel, David (2011): Evidence for early morphological decomposition: combining masked priming with magnetoencephalography. Journal of Cognitive Neuroscience, 23(11), pp. 3366–3379. Libben, Gary (1998): Semantic transparency in the processing of compounds: consequences for representation, processing, and impairment. Brain and Language, 61(1), pp. 30–44. Libben, Gary (2003): Morphological parsing and morphological structure. In: Egbert M.H. Assink — Dominiek Sandra (eds.), Reading Complex Words: Cross-Language Studies. Amsterdam: Kluwer, pp. 221–239. Longtin, Catherine-Marie — Segui, Juan — Hallé, Pierre A. (2003): Morphological priming without morphological relationship. Language and Cognitive Processes, 18(3), pp. 313–334. Marslen-Wilson, William D. — Tyler, Lorraine Komisarjevsky — Waksler, Rachelle — Older, Lianne (1994): Morphology and meaning in the English mental lexicon. Psychological Review, 101(1), pp. 3–33. Maslov, Jurij Sergeevič (1963): Morfologija glagol�nogo vida v sovremennom bolgarskom literaturnom jazyke. Moskva — Leningrad: Izdatel�stvo akademii nauk SSSR. Matthews, Peter Hugoe (1972): Inflectional Morphology: A Theoretical Study Based on Aspects of Latin Verb Conjugation. Cambridge: Cambridge University Press. Matthews, Peter Hugoe (1991): Morphology. Cambridge: Cambridge University Press. McQueen, James — Cutler, Anne (1998): Morphology in word recognition. In: Andrew Spencer — Arnold M. Zwicky (eds.), The Handbook of Morphology. Oxford — Malden, MA: Blackwell. Monsell, Stephen (1985): Repetition and the lexicon. In: Andrew W. Ellis (ed.), Progress in the Psychology of Language: Vol. 2. London: Lawrence Erlbaum Associates, pp. 147–195. Münte, Thomas F. — Say, Tessa — Clahsen, Harald — Schiltz, Kolja — Kutas, Marta (1999): Decomposition of morphologically complex words in English: evidence from event-related brain potentials. Cognitive Brain Research, 7(3), pp. 241–253. Newman, Aaron J. — Ullman, Michael T. — Pancheva, Roumyana — Waligura, Diane L. — Neville, Helen J. (2007): An ERP study of regular and irregular English past tense inflection. NeuroImage, 34(1), pp. 435–445. Nikolova, Rossitza — Jarema, Gonia (2002): Interaction of morphological structure and prefix transparency in the processing of Bulgarian aspectual verb forms. Brain and Language, 81(1–3), pp. 649–665. 70 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2015 Peciar, Štefan (1966): Funkcie slovesnej predpony uv slovenčine. Slovenská reč, 31(5), pp. 261–269. Perlmutter, David (1988): The Split Morphology Hypothesis: evidence from Yiddish. In: Michael Hammond — Michael Noonan (eds.), Theoretical Morphology: Approaches in Modern Linguistics. San Diego, CA, et al.: Academic Press, pp. 79–100. Pinker, Steven (1997): Words and rules in the human brain. Nature, 387, pp. 547–548. Pinker, Steven (1998): Words and rules. Lingua, 106(1–4), pp. 219–242. Pinker, Steven (1999): Words and Rules: The Ingredients of Language. New York, NY: Basic Books. Pinker, Steven — Ullman, Michael T. (2002): The past and future of past tense. Trends in Cognitive Sciences, 6(11), pp. 456–463. Plaut, David C. — Gonnerman, Laura M. (2000): Are non-semantic morphological effects incompatible with a distributed connectionist approach to lexical processing? Language and Cognitive Processes, 15(4–5), pp. 445–485. Rastle, Kathleen — Davis, Matt H. — Marslen-Wilson, William D. — Tyler, Lorraine Komisarjevsky (2000): Morphological and semantic effects in visual word recognition: a time-course study. Language and Cognitive Processes, 15(4–5), pp. 507–537. Rastle, Kathleen — Davis, Matt H. — New, Boris (2004): The broth in my brother’s brothel: morpho-orthographic segmentation in visual word recognition. Psychonomic Bulletin and Review, 11(6), pp. 1090–1098. Rueckl, Jay G. — Mikolinski, Michelle — Raveh, Michal — Miner, Caroline S. — Mars, Frank (1997): Morphological priming, fragment completion, and connectionist networks. Journal of Memory and Language, 36(3), pp. 382–405. Sandra, Dominiek (1990): On the representation and processing of compound words: automatic access to constituent morphemes does not occur. Quarterly Journal of Experimental Psychology: Section A: Human Experimental Psychology, 42(3), pp. 529–567. Scalise, Sergio (1984): Generative Morphology [Studies in Generative Grammar, 18]. Dordrecht: Foris. Scalise, Sergio (1988): Inflection and derivation. Linguistics, 26(4), pp. 561–582. Schriefers, Herbert — Friederici, Angela D. — Graetz, Patty A. M. (1992): Inflectional and derivational morphology in the mental lexicon: symmetries and asymmetries in repetition priming. Quarterly Journal of Experimental Psychology: Section A: Human Experimental Psychology, 44(2), pp. 373–390. Schneider, Walter — Eschman, Amy — Zuccolotto, Anthony P. (2002): E-Prime, v. 1.2 [software]. Pittsburgh, PA: Psychology Software Tools Inc. Schreuder, Robert — Baayen, R. Harald (1994): Prefix stripping re-revisited. Journal of Memory and Language, 33(3), pp. 357–375. Schreuder, Robert — Baayen, R. Harald (1995): Modeling morphological processing. In: Laurie B. Feldman (ed.), Morphological Aspects of Language Processing. Hillsdale, NJ: Lawrence Erlbaum Associates, pp. 131–154. Schreuder, Robert — Baayen, R. Harald (1997): How complex simplex words can be. Journal of Memory and Language, 37(1), pp. 118–139. Slabakova, Roumyana (2001): Aspectual constraints in the mental lexicon. Journal of Psycholinguistic Research, 30(2), pp. 197–217. Šlosar, Dušan (1981): Slovotvorný vývoj českého slovesa. Brno: Univerzita J. E. Purkyně. Šlosar, Dušan (1986): Slovotvorba. In: Arnošt Lamprecht — Dušan Šlosar — Jaroslav Bauer, Historická mluvnice češtiny. Praha: Státní pedagogické nakladatelství, pp. 251–340. Smith, Philip T. — Sterling, Christopher M. (1982): Factors affecting the perceived morphemic structure of written words. Journal of Verbal Learning and Verbal Behavior, 21, pp. 704–721. Solomyak, Olla — Marantz, Alec (2010): Evidence for early morphological decomposition in visual word recognition. RADKA JULíNKOVá — DENISA BORDAG 71 Journal of Cognitive Neuroscience, 22(9), pp. 2042–2057. Sonnenstuhl, Ingrid — Eisenbeiss, Sonja — Clahsen, Harald (1999): Morphological priming and the mental lexicon: evidence from German. Cognition, 72, pp. 203–236. Stanners, Robert F. — Neiser, James J. — Hernon, William P. — Hall, Roger (1979): Memory representation for morphologically related words. Journal of Verbal Learning and Verbal Behavior, 18(4), pp. 399–412. Taft, Marcus (1979): Recognition of affixed words and the word frequency effect. Memory and Cognition, 7(4), pp. 263–272. Taft, Marcus (1981): Prefix stripping revisited. Journal of Verbal Learning and Verbal Behavior, 20, pp. 289–297. Taft, Marcus (2004): Morphological decomposition and the reverse base frequency effect. Quarterly Journal of Experimental Psychology: Section A: Human Experimental Psychology, 57(4), pp. 745–765. Taft, Marcus — Forster, Kenneth I. (1976): Lexical storage and retrieval of polymorphemic and polysyllabic words. Journal of Verbal Learning and Verbal Behavior, 15(6), pp. 607–620. Uher, František (1987): Slovesné předpony. Brno: Univerzita J. E. Purkyně. Ullman, Michael T. (2001): A neurocognitive perspective on language: the declarative/ procedural model. Nature Reviews Neuroscience, 2(10), pp. 717–726. Ullman, Michael T. (2004): Contributions of memory circuits to language: the declarative/procedural model. Cognition, 92(1–2), pp. 231–270. Ullman, Michael T. — Estabrooke, Ivy V. — Steinhauer, Karsten — Brovetto, Claudia — Pancheva, Roumyana — Ozawa, Kaori — Mordecai, Kristen — Maki, Pauline (2002): Sex differences in the neurocognition of language. Brain and Language, 83, pp. 141–143. Vannest, Jennifer — Boland, Julie E. (1999): Lexical morphology and lexical access. Brain and Language, 68(1–2), pp. 324–332. Veselý, Luboš (2010): Ke slovesnému vidu v češtině. Naše řeč, 93(3), pp. 113–124. Wurm, Lee H. (1997): Auditory processing of prefixed English words is both continuous and decompositional. Journal of Memory and Language, 37(3), pp. 438–461. Zwitserlood, Pienie (1994): The role of semantic transparency in the processing and representation of Dutch compounds. Language and Cognitive Processes, 9(3), pp. 341–368. 72 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2015 APPENDIX Target Prime dřu dřu dřeš dříč kopu kopu kopeš kopáč kreslím kreslím kreslíš kreslíř lezu lezu lezeš lezec piju piju piješ pijan plavu plavu plaveš plavec rvu rvu rveš rváč řečním řečním řečníš řečník štvu štvu štveš štváč zabiju zabiju zabiješ zabiják zvoním zvoním zvoníš zvoník žebrám žebrám žebráš žebrák cíl cíl cílem cílíš čich čich čichem čicháš dřep dřep dřepem dřepíš hlt hlt hltem hltáš hmat hmat hmatem hmatáš klam klam klamem klameš kmit kmit kmitem kmitáš kouř kouř kouřem kouříš kov kov kovem kováš lov lov lovem lovíš obal obal obalem obalíš pláč pláč pláčem pláčeš vrt vrt vrtem vrtáš zápas zápas zápasem zápasíš zob zob zobem zobeš experiment 1: Materials. RADKA JULíNKOVá — DENISA BORDAG 73 Inflection Target uTarget zucítí cítí zbledne bledne udělá dělá zbohatne bohatne uhlídá hlídá zboří boří uloví loví zdědí dědí umlátí mlátí zhatí hatí upeče peče zhřeší hřeší uslyší slyší zhubne hubne usmaží smaží zkazí kazí uspoří spoří zkropí kropí uškodí škodí zkrotí krotí ušpiní špiní zláká láká utrápí trápí zmaří maří uvaří vaří zmate mate uvěří věří změří měří uvidí vidí zmrzne mrzne Transparent derivation Target uTarget zubrousí brousí zbourá bourá udrží drží zbrousí brousí udupe dupe zbrzdí brzdí ukolébá kolébá zdeptá deptá ukončí končí zhasne hasne ukotví kotví zhodnotí hodnotí umaže maže zchátrá chátrá umyje myje zchladí chladí upálí pálí zkrátí krátí uřeže řeže zmešká mešká ušlape šlape zmírní mírní utají tají zmrazí mrazí utlačí tlačí znásobí násobí utlumí tlumí zpracuje pracuje utvoří tvoří ztlumí tlumí 74 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2015 Opaque derivation Target uTarget zudělí dělí zdraží draží uchová chová zdrhne drhne ukáže káže zdrží drží ukryje kryje zchytá chytá uplatí platí zjedná jedná upraví praví zjistí jistí usoudí soudí zklame klame uspěje spěje zkreslí kreslí uteče teče zmíní míní uvalí valí zmydlí mydlí uváží váží zradí radí uvede vede zřídí řídí uvrhne vrhne ztrhá trhá uzdraví zdraví zvětrá větrá uzná zná zvládne vládne Non-prefixed verbs Target uTarget zubytuje zdobí udeří zdolá udolá zdvihá uhodne zkouší uhrane zlatí umístí zlobí unaví zmatkuje uniká zmítá určí zmlkne urguje značí usiluje zpívá uskupí zrcadlí uspíší zvedá uštkne zvoní utuží zvučí experiment 2: Materials. RADKA JULíNKOVá — DENISA BORDAG 75 ABSTRAKT: Zpracování a reprezentace různých typů českých afixů. Studie se zabývá zpracováním morfologicky komplexních slov v češtině. V experimentu 1 využíváme morfologický repetition priming k ověření tzv. Split Morphology Hypothesis, tj. zjišťujeme, zda jsou slovní formy vzniklé derivací vs. flektivně v českém mentálním lexikonu uloženy stejným, či odlišným způsobem. Výsledky ukazují, že primingové efekty jsou podstatně silnější v případě flektivně vzniklých forem než v případě forem derivovaných, což svědčí o odlišném zpracování flexe a derivace v češtině; zatímco flektivně vzniklé formy jsou během zpracování rozloženy úplně, derivované formy nejsou rozloženy buď vůbec, nebo jen částečně. V experimentu 2 se zabýváme dvěma výzkumnými otázkami. Zaprvé testujeme psycholingvistickou realitu lingvistické distinkce mezi dvěma typy slovesných prefixů: (a) prefixy „čistě“ vidovými (které z nedokonavého slovesa činí sloveso dokonavé, srov. např. hřešit — zhřešit) a (b) prefixy derivačními (např. krátit — zkrátit). Výsledky neposkytují pro tuto distinkci žádné psycholingvistické důkazy. Zadruhé zkoumáme, jakou roli má při zpracování jazyka sémantická transparentnost. Experiment přináší důkazy o tom, že netransparentní derivovaná slovesa jsou zpracovávána pomaleji, což je s největší pravděpodobností způsobeno jejich dvojím vyhledáváním / dvojí reanalýzou. Radka Julínková | Department of General Linguistics, Faculty of Arts, Palacký University in Olomouc <[email protected]> Denisa Bordag | Herder-Institut, University of Leipzig <[email protected]>