etusivulle

FiCLAn syyssymposiumi

Prosessointi ja kokemus

24.–25. syyskuuta 2004
Joensuun yliopistossa
salissa Agora H1

Suomen kognitiivisen kielentutkimuksen yhdistys FiCLA ja Joensuun yliopiston kieliaineet järjestävät yhteisen symposiumin, joka on tarkoitettu kaikille kielen ja mielen kognitiivisesta tutkimistavasta kiinnostuneille opiskelijoille ja opettajille.

Ohjelma

Perjantai 24.9.

11.00 Jussi Niemi (Joensuun yliopisto)& Oili Karihalme (FiCLA, Turun yliopisto):
Tervetuliaissanat / Welcoming words
11.15 Plenaariesitelmä: Chris Sinha (University of Portsmouth): Cultural experience, embodiment, language, and action
12.15 Timo Honkela (Teknillinen korkeakoulu): Ideas on Language Use and Motivation

13.00 Lounastauko

14.00 Plenaariesitelmä: Esa Itkonen (Turun yliopisto): What can cognitive linguistics learn from typological linguistics? (abstract)
15.00 Heli Tissari (Helsingin yliopisto): Conceptualising the English noun PRIDE: Some corpus-based remarks (abstract)

15.30 Kahvitauko

16.00 Marjatta Lehtinen (Joensuun yliopisto): How words become acts: On the construction of emotions in the speech by women in a self-help therapy group (abstract)
16.30 Viljo K. Martikainen (CSC): Concepts and the processual mental in the RA-model of man (abstract)

17.00 – n. 18.00 Yleiskeskustelua

20.00 Illallinen kaupungilla (ilmoittautumiset Tero Kainlaurille, kainlaur@cc.joensuu.fi, viimeistään 20.9., omakustanteinen).

Lauantai 25.9.

9.00 Plenaariesitelmä: Gabriella Vigliocco (University College London): Representing the meaning of object and action words: The Featural and Unitary Semantic Space (FUSS) hypothesis (abstract)
10.00 Jussi Niemi & Sinikka Niemi (Joensuun yliopisto): Biological Age of Speakers and/or Frequency of Lexical Types as Determinants of Lexical Performance (abstract)
11.00 Alexandre Nikolaev (Joensuun yliopisto): Alku- ja loppusegmentin vaikutus suomen nominien prosessointiin (abstract)

11.30 Lounastauko

12.30 Plenaariesitelmä: Jordan Zlatev (Lundin & Uumajan yliopisto): The dependence of language on consciousness (abstract)
13.30 Seija Pekkala (Helsingin yliopisto): Interpretation of the errors made by Alzheimer´s patients in the semantic fluency task (abstract)
14.00 Johanna Salonen (Turun yliopisto), Arto Mustajoki (Helsingin yliopisto), Marie Cheour (University of Miami) & Elina Pihko (HUCH ja HBRC): N400 in Fluent Finnish-Russian Bilinguals: Comparison of EEG and MEG Findings (abstract)

14.30 Loppusanat / Closing words

Ilmoittautuminen ja maksut

FiCLA:n jäsenmaksun maksaneille tilaisuus on ilmainen, samoin Joensuun yliopiston kieliaineiden perustutkintojen opiskelijoille. FiCLA:n jäsenmaksun maksamiseksi yhteyden voi ottaa Tero Kainlauriin (kainlaur@cc.joensuu.fi). Maksavien osallistujien ja illalliselle tulijoiden toivotaan ilmoittautuvan viimeistään 20.9. Kainlaurille. Yleinen osallistumismaksu on 20 euroa, opiskelijoilta 10 euroa. FiCLA:n tilinumero on Sampo 800012-70761654.

Ohjelman aikojen pienet muutokset ovat mahdollisia.

TERVETULOA JOENSUUHUN PUHUMAAN JA KUUNTELEMAAN!

Abstracts

Marjatta Lehtinen (marjatta.lehtinen@joyx.joensuu.fi):
How words become acts: On the construction of emotions in the speech by women in a self-help therapy group

According to Bamberg (1997: 209), studies on language and affect tend to focus on the communicative aspect of language use, examining how affective signals are interpreted by the recipient; this perspective sets the focus on the descriptive function of language. Another approach is to investigate how linguistic forms reflect and construct emotions; this view takes the speaking subject as the point of departure. The focus is on the expressive rather than descriptive function of language. As Bamberg points out (1997: 209), the communicative approach, i.e., the listener’s perspective, treats the two systems - expressing emotions and expressing linguistic meanings - as separate, though parallel, devices in use. However, from the perspective of the speaking subject this is implausible. The speaker does not produce the expression, the linguistic meaning, as isolated from the feeling; uttering the linguistic expression is emotional action (cf. Schafer 1976; Hintikka 1999: 82). My paper deals with the construction of emotions in an act of unburdening the self, in which the speaker uses affect in a specific therapy context. The setting is the interaction carried out by a feminist therapy group applying the method of FRT (Feminist Radical Therapy). FRT as therapy belongs to autonomous (self-help) group therapies. In FRT-sessions women meet regularly to work with and solve problems of their lives (on the concept FRT, see Hintikka 1999, 6-7). The FRT-method relates to assertion training, in which language is set for teaching people to express themselves directly, and owning their feelings (Gallois 1994; Hintikka 1999: 82). In general, in therapeutic discourse, it matters how freely the speaker communicates; therapy is done with words. FRT lacks the asymmetry between the therapist and the client characterizing mainstream therapy settings (cf. Labov and Fanshel 1977). In working with a problem, the speaker explores his/her unconscious motives; successful communication means direct expression, getting in touch with one’s repressed feelings. The FRT-self-expression method enables the speaker to work with her repressed emotions by re-experiencing them (verbally and non-verbally) via regression; working them through leads to gaining more independence, to use the words by Hintikka (1999), in the therapeutic event, ‘words become acts’. I shall outline some basic features of the language used in FRT, and in particular, describe some ways of self-expression in a speech event called counselling talk, illustrating it by means of an excerpt of a dialogue.

References:
Bamberg, M. 1997. Emotion talk(s): The ro!e of perspective in the construction of emotions. Niemeier S. & Dirven R. (eds), The Language of Emotions. Amsterdam: John Benjamins. 209-225.

Ernst, S. & Goodison L.1988 (1981). In Our Own Hands. A Book of Self-help Therapy. London: The Women's Press.

Gallois, C. 1994. Group membership, social rules,and power. A social- psychological perspective on emotional communication. Journal of Pragmatics 22. 301-324.

Hintikka, S. 1999. Feministinen Radikaaliterapia itsehoitomenetelmänä ja naisen identiteetin tukijana. Sosiaalipsykologian pro gradu-tutkielma. Tampereen yliopisto.

Labov, W. & Fanshel D. 1977. Therapeutic Discourse. Psychotherapy as Conversation. New York: Academic Press.

Lehtinen, M. 1997. Say how You Feel! On Feminist Therapeutic Discourse. Erikoiskielet ja käännösteoria. VAKKI Symposium XVII, Vöyri, 22.-23.2.1997. 173 - 181.

Schafer, R. 1976. A New Language for Psychoanalysis. New Haven: Yale University Press.

Niemi, Jussi and Sinikka Niemi (jussi.niemi@joensuu.fi, sinikka.niemi@joensuu.fi)
Biological Age of Speakers and/or Frequency of Lexical Types as Determinants of Lexical Performance

Recurrent patterns of complex morphological (inflectional) relationships like those pertaining to English irregular tense forms (cling : clung, string : strung, hence also bring : *brung) tend to create analogical, word_shape_based, schematic links in the morpholexical network. The present study derives its data from two languages in which the selection of affixes is subject to specific morpho(phono)logical constraints, viz., Finnish and Swedish. The subjects of the wug experiments (Berko 1958) with pseudowords are normal speakers of the 7 to 15 year age range (as well as, in Finnish, four members of an extended family with three generations of attested Specific/Familial Language Impairment (SLI/FLI)).

The present results, based on analyses for Niemi and Hugdahl (submitted) and on re-analyses of Niemi and Niemi (2002), show, inter alia, (a) that the morphologies of normal children undergo considerable development still at the present age range, implying the emergent, slowly stabilizing nature during the ontogeny of complex grammatical structures. (b) Moreover, the phonologically conditioned (rule_driven) inflectional relationships are mastered better than the lexically conditioned ones. (c) In the calculations for the Finnish data we observed that in the normal speakers the frequency of use correlates positively with lexical performance while the language-impaired speakers show no such frequency effect.

In our presentation we will discuss the interpretations and implications of the empirical data.

References:
Berko, J. (1958). The child's learning of English morphology. Word 14: 150_177.

Niemi, J. & K. Hugdahl (submitted). Plural morphology in normal acquisition and in Familial Language Impairment: Experimental observations on phonologically and lexically conditioned paradigms. Ms. submitted February 2004.

Niemi, S. & J. Niemi 2002. Emergent nature of morphological paradigms: Plural inflection in Swedish and Finnish. In: Renate Pajusalu & Tiit Hennoste (eds), Tähendusepüüdja/Catcher of the Meaning. A Festschrift to Haldur Õim on the Occasion of His 60th Birthday. Publications from the Department of General Linguistics, 3. University of Tartu. Pp. 285-296.

Alexandre Nikolaev:
Alku- ja loppusegmentin vaikutus suomen nominien prosessointiin

Yksi lingvistiikassa usein käytetyistä testimuodoista on ns. WUG-testi (ks. Berko 1958), jossa käytetään kielen morfofonologisia lainalaisuuksia vastaavia epäsanoja. Epäsanojen käyttö mahdollistaa leksikaalisen efektin välttämisen lisäksi kielen tietyn osa-alueen produktiivisuustendenssien kartoittamisen (näin epäsanat voidaan rinnastaa neologismeihin). Neologismien tapaan suomen kielen morfofonologisia lainalaisuuksia vastaava sana sijoittuu tiettyyn taivutustyyppiin. Usein epäsanat muodostetaan oikeiden sanojen pohjalta (jotka kuuluvat jo tiettyyn taivutustyyppiin) muuntelemalla esim. alkusegmenttiä. Kuinka paljon epäsanan täytyy erota aidosta sanasta, jotta se ei assosioidu enää tähän vaan sijoittuu neologismien tapaan produktiiviseen taivutusparadigmaan? Vaikuttaako epäsanan prosessointiin se, mikä segmentti aidossa sanassa on vaihdettu tai poistettu? Miten voimme suhteuttaa epäsanojen prosessoinnin yhteydessä tehtyjä päätelmiä aitojen suomen kielen sanojen prosessointiin?

Vastatakseni näihin kysymyksiin, suoritin kokeita, joissa käytin samalla algoritmilla muunnettuja epäsanoja. Kaikki epäsanat oli muodostettu kahden kilpailijaparadigman pohjalta, joista toista edustaa patsas:patsaan -taivutustyyppi ja toista keskus:keskuksen -taivutustyyppi. Molemmissa on yksi konstantti aines, joka on leimallinen juuri kyseiselle taivutustyypille, nimittäin taivutustyypissä patsas tällaisena konstanssina on loppusegmentti -As (leksikaalinen frekvenssi 95,6%) ja vatsaavasti taivutustyypissä keskus -Us (l.f. - 89,2%). Epäsanat on muodostettu seuraavaa algoritmia käyttäen: 1. muutetaan alkukonsonantti / alkuvokaali toiseksi; 2. poistetaan alkukonsonantti / lisätään alkuvokaalin eteen konsonantti; 3. vaihdetaan loppusegmentti oppositiossa olevan ryhmän prototyyppiseen loppusegmenttiin. Tällä muunnoksella on saatu kolme epäsanaryhmää kustakin taivutustyypistä (yhteensä siis 60 epäsanaa). Kokeeseen osallistui 46 14-vuotiasta Joensuun Normaalikoulun oppilasta. Koe oli suullinen ja jokaisen koehenkilön piti taivuttaa 60 epäsanaa yksikön genetiivissä (Minä otan …).

Testitulosten analyysi johti seuraaviin päätelmiin:

– alkusegmentin muunnos ei vaikuta ratkaisevasti epäsanojen prosessointiin (muunnoksella tarkoitetaan alkukonsonantin poistoa / vaihtoa toiseen; konsonantin lisäystä alkuvokaalin eteen sekä alkuvokaalin vaihtoa)

– loppusegmentti ohjaa taivutusta kaikissa tarkasteltavissa ryhmissä; loppusegmentin vaihto prototyyppiseen kilpailijaparadigman loppusegmenttiin johtaa siihen, että epäsana taivutetaan kilpailijaparadigman mukaisesti useammin kuin toisen ryhmän epäsana (ryhmien välillä on tilastollisesti merkitsevä ero)

– kolmitavuinen epäsana säilyttää sen alla olevan oikean sanan ”latauksen” paremmin kuin kaksitavuinen epäsana (erolla on tilastollista merkitystä)

– epäsanojen taivutuksen ja sen alla olevan oikean sanan käyttöyleisyyden välillä ei ole merkitsevää korrelaatiota

Tuloksena voidaan todeta, ettei tiettyyn paradigmaan sijoittuminen johdu niinkään tietystä segmentistä (tai sen paikasta), vaan pikemmin siitä, onko tämä segmentti sen paradigman prototyyppinen edustaja. Mikäli sanan segmenteistä yksikään ei ole jonkin paradigman prototyyppinen edustaja, niin muuntelemalla jotakin segmenttiä vapautamme saadun epäsanan aidon sanan paradigmasta, minkä jälkeen produktiivinen taivutustyyppi vetää sen puoleensa.

Johanna Salonen, Arto Mustajoki, Marie Cheour & Elina Pihko:
N400 in Fluent Finnish-Russian Bilinguals: Comparison of EEG and MEG Findings

N400 is an event-related potential reflecting semantic processing and contextual congruency effect. It is a long lasting response at 350-600 msec after stimulus onset. Incongruent sentence-ending words (e.g. ’There is too much salt in the computer’) elicit more negative responses than those final words that are semantically coherent with the rest of the sentence (e.g. ’There is too much salt in the food’). Previous electroencephalography studies show that, in bilinguals, the latency of N400 is delayed in their second language in comparison with their native language and the strength of activation is smaller.

We measured seven volunteers with a 60-channel electroencephalography (EEG) and 306-channel magnetoencephalography (MEG) at BioMag-laboratory (Helsinki University Central Hospital). All the subjects were Finnish natives who were fluent in Russian but had learned it in adulthood. They sat in a magnetically shielded room and read 100 Finnish and 94 Russian sentences. The final word of the sentences was either semantically expected and congruent or surprising and incongruent. The total number of sentences was thus 388. A clear N400 response was seen in EEG data. The responses to incongruent final words were more negative than the responses to congruent words both in Finnish and in Russian. Also a difference between the languages was observed. The responses to Russian stimuli were smaller both in the congruent and incongruent condition than the responses to Finnish sentences. Also the latency of responses to Russian words was delayed in comparison with the responses to Finnish stimuli. These findings are in accordance with previous EEG studies. The peak amplitude was delayed by 96 msec. The processing of Russian ended 111 msec later than processisng of Finnish. No differences in the onset latency of processing were observed, but a tendency to process Russian final words longer than Finnish ones was seen. Also in the MEG data, a clear N400 response was detected both in Finnish and in Russian, and just like in the EEG data, the responses to incongruent words were larger than responses to congruent stimuli. The amplitude difference between Finnish and Russian stimuli detected by EEG was not seen in the MEG responses. Instead, like in EEG data, the peak amplitude was elicited 130 msec later in Russian than in Finnish. The differences in EEG and MEG responses can be explained at least partially by the fact that MEG detects only part of the activation in the brain that EEG measures. Also the influence of different alphabets and reading speed has to included into the interpretation of our results, although on these matters not many studies exist.

Seija Pekkala:
Interpretation of the errors made by Alzheimer’s patients in the semantic fluency task

The poor performance of the patients with Alzheimer’s disease (AD) on semantic tasks (e.g., naming, matching, and feature decision tasks) is usually interpreted as purely a semantic impairment that may result from deteriorated perceptual functions, loss of the semantic information, impaired access to the semantic information, or a combination of the last two causes.

Some of the current connectionist models assume that damage to the semantic features and/or their connections, noise-induction in the speech processing system, and changes in the rate at which activated information decays (Dell 1986; Farah & McClelland 1991; Martin et al. 1994; Gonnerman et al. 1997; Devlin et al. 1998; Harley 1998) may interfere with the performance on the semantic tasks (Harley 1998; Moss et al. 2002). Nevertheless, as far as word production is concerned, damaged or lost features may affect the functioning at the semantic level of the mental lexicon, as well as the lemma (lexical) and the phonological level, due to the interconnectivity between the levels (Harley 1998). The findings (e.g., intrusions and perseverations) obtained from AD patients’ performances on the semantic fluency task, in which words were produced according to some semantic constraint (e.g., animals, vegetables, and verbs denoting playing sports), indicate a need to discuss their appearance not only in relation to the semantic stage but also to these other stages of the word production process (Pekkala, 2004). However, the interpretation of the loci of errors seems to depend on theories.

REFERENCES:
Dell, G. S. (1986). A spreading-activation theory of retrieval in sentence production. Psychological Review, 93, 283-321.

Devlin, J. T., Gonnerman, L. M., Andersen, E. S., & Seidenberg, M. S. (1998). Category-specific semantic deficits in focal and widespread brain damage: A computational account. Journal of Cognitive Neuroscience, 10, 77-94.

Farah, M. J., & McClelland, J. L. (1991). A computational model of semantic memory impairment: Modality specificity and emergent category specificity. Journal of Experimental Psychology: General, 120, 339-357.

Gonnerman, L. M., Andersen, E. S., Devlin, J. T., Kempler, D., & Seidenberg, M. S. (1997). Double dissociation of semantic categories in Alzheimer’s disease. Brain and Language, 57, 254-279.

Harley, T. A. (1998). The semantic deficit in dementia: Connectionist approaches to what goes wrong in picture naming. Aphasiology, 12, 299-318.

Martin, N., Dell, G. S., Saffran, E. M., & Schwartz, M. F. (1994). Origins of paraphasias in deep dysphasia: Testing the consequences of a decay impairment to an interactive spreading activation model of lexical retrieval. Brain and Language 47, 609-660.

Moss, H. E., Tyler, L. K., & Devlin, J. T. (2002). The emergence of category-specific deficits in a distributed semantic system. In E.M.F. Forde & G.W. Humphreys (Eds.), Category specificity in brain and mind (pp. 115-147). Hove, UK: Psychology Press.

Pekkala, S. (2004). Semantic fluency in mild and moderate Alzheimer’s disease. Doctoral dissertation. Publications of the Department of Phonetics, 47. Helsinki: University of Helsinki. http://ethesis.helsinki.fi.

Viljo K. Martikainen (viljo.martikainen@csc.fi):
CONCEPTS and the processual mental in the RA-model of man

In my dissertation work (Martikainen 2004) I have created a system model of Man and his concepts. They are built of real and testable memory representations of entities of the being. The totality of the human mental can be structured and its functions understood by seeing it as a continuum of mental states caused by the recalled memory representations of the object entities in subject's attention. Usually the active memory representations are situation relevant set of concepts and their attributes explaining the sensed, thought, or imagined features of subject's actual environment.

The RA-model is based on a set of findings of sciences: 1. Man is a rational, social, and biological actor created by the evolutionary and cultural processes of human kind. 2. The dorsal and ventral streams of visual information flows in our brains - found by neuropsychology - exemplify the on-line and memory-line processes explaining man's interactions with his environment. 3. Man's testable memory representations and their situation relevant recallings into our working memory form the empirical basis of our concepts and their functions as tools of thinking and understanding.

Human concepts as structures of man's memory representations have the base-, time-, and theory lines in their system structure. Signs, words, sentences, and all other linguistic expressions form parts of the sensory attributes on the base line of our concepts. All sensory attributes can function as search attributes recalling the other situation relevant concepts and their sensory and structural attributes to explain the sensed or thought entity in the focus of our attention.

Our concepts function as situation relevant and handy tools for thinking, planning, and decision-making. RA's concepts enable thus a natural explanation of the human mental. Concepts can be used to disclose the mental dimensions of cultures, which seem to be hidden in the values, norms, language, signs, linguistic expressions, and action models adopted by the members of the culture. Thus, they serve also the conscious control of man's mental and materially manifesting actions and activities.
RA-model of man is a meta-model, which enables to see the basic problems of human sciences and philosophy from a new and natural point of view. Our linguistic ability seems to be based on the processual formation of concepts and their sensory and structural attributes. Thus signs, words, sentences, and other linguistic expressions - used in human communicative acts - function as sensory attributes of concepts of entities. The genesis of signs and linguistic expressions happentities processually in the situations they are used, as Wittgenstein claimed. Their referential power can be explained by their functioning as search attributes recalling the situation relevant concepts to subject's working memory to enable a conscious mental state to clarify the sensed or thought.

The RA-model of man and his mental seems to offer a natural frame to review the bordelien between the conscious and processual in the human linguistic abilities. The SCENE OF EXPERIENCING AND STEERING in RA-model shows this bordeline and explaines how all our speech and other motor actions are steered by a situation relevant combination of the conscious and processual steering.

References

Martikainen, Viljo. K. (2004) Concepts and Mind as Dynamic Memory Systems Structuring the Human Mental. Helsinki: JOKO-VM Oy.

Heli Tissari (heli.tissari@helsinki.fi):
Conceptualising the English word PRIDE: Some corpus-based remarks

This paper presents a relatively small-scale historical study of how well Kövecses’s (1986: 39–60) prototype for the concept of ‘pride’ agrees with English language data on the word PRIDE (noun, verb and adjective). The data comes from five English language corpora which cover a period from the beginning of the 15th century to the beginning of the 1990s. The method is a close reading of PRIDE items in an 8-line-context. The questions asked are: (1) What causes pride? (2) Which related concepts are mentioned in the context? (3) Which behavioural reactions are mentioned in the context? (4) What kind of conceptual metaphors occur nearby the items? Questions (2) and (4) relate to each other, because they both not only reveal to us what kind of other concepts are associated with the concept of ‘pride’, but also how pride is evaluated: is it considered good or bad to be proud? Special attention is also paid to the concept of ‘justified pride’, as introduced by Kövecses (1986: 46–47). The paper ends with some remarks on the usefulness of corpus data for cognitive studies, on modelling concepts, on cognitive studies and norms of behaviour, and on questions for further research. This study is part of a larger project on the meaning and development of English emotion words since the Renaissance (Heikkinen & Tissari 2002, Tissari 2003 & 2004).

Primary reference:

Kövecses, Z. (1986) Metaphors of anger, pride and love. (Pragmatics & beyond VII: 8.) Amsterdam/Philadelphia: John Benjamins.

Other references:

Heikkinen, K. & Tissari, H. (2002) “GEFEOH and GEBLISSA or happy birthday! On Old English BLISS and Modern English HAPPY.” Variation past and present: VARIENG studies on English for Terttu Nevalainen (Mémoires de la Societé Néophilologique de Helsinki LXI), ed. by Helena Raumolin-Brunberg, Minna Nevala, Arja Nurmi & Matti Rissanen. Helsinki: Société Néophilologique. 59–76.

Tissari, H. (2003) LOVEscapes: Changes in prototypical senses and cognitive metaphors since 1500. (Mémoires de la Société Néophilologique de Helsinki LXII.) Helsinki: Société Néophilologique.

Tissari, H. (2004) “Vistas of fear and hope: On metaphors occurring with the English words FEAR and HOPE.” Papers from the 30th Finnish Conference of Linguistics Joensuu, May 15–16, 2003, ed. by Marja Nenonen. Joensuu: University of Joensuu. 214–220.

Gabriella Vigliocco:
On concepts, words and syntax: the featural and unitary semantic space (FUSS) hypothesis and beyond

In 1910, John Dewey wrote: "To grasp the meaning of a thing, an event, or a situation is to see it in its relations to other things; to note how it operates or functions, what consequences follow from it, what causes it, what uses it can be put to". In this passage, he captures two fundamental questions in the investigation of meaning: how are meanings related? How is each meaning structured. In the talk I will address these two interrelated questions. I will start by introducing the Featural and Unitary Semantic Space (FUSS) hypothesis of the representation of the meanings of words referring to objects and events. FUSS integrates a number of previous claims and research strands encompassing psycholinguistics, cognitive psychology and neuroscience into a hypothesis which we have implemented into a statistical model. I will then turn to use FUSS as a tool to explore the representation and processing of other linguistic information such as grammatical class, which is correlated with meaning distinctions.

Jordan Zlatev (Jordan.Zlatev@ling.lu.se):
The dependence of language on consciousness

It is nearly universally accepted, among generative and cognitive linguists alike, that language is learned, represented and processed (almost exclusively) “unconsciously”. I wish to challenge this dogma on the basis of both conceptual and empirical evidence.Our knowledge of language needs to be accessible to consciousness to be able to account for its central property of normativity (Wittgenstein, Itkonen). Furthermore, since linguistic knowledge is “knowledge” (rather than habits, associations or connections), it must be at least in principle accessible to consciousness according to Searle’s “connection principle”. More empirically, I turn to work in language acquisition, showing that far from being “acquired unconsciously” through generativist or connectionist "devices", the learning of a language by both children and adults is based on various forms of analysis, presupposing conscious thought (Mandler, Tomasello, Bloom). Finally, while language use is indeed largely automated and proceduralized, it is nevertheless supervised by consciousness as shown in self-corrections and discourse coherence (Chafe, Donald). My major claim is that only by a better understanding of consciousness and its integration in the study of language will we be able to make a breakthrough in our understanding of the latter.

Esa Itkonen:
What can cognitive linguistics learn from typological linguistics?

Typological linguistics investigates simultaneously the diversity and the unity of the world’s languages. As was noted by Hymes in 1964, this type of investigation (whether it is called ‘typological’ or ‘anthropological’) is threatened by two complementary types of failure. On the one hand, there is FAILURE OF EMPATHY, “the ‘classical’ error, too readily assuming a single norm, often our own, by which to judge diversity”. On the other hand, there is FAILURE OF IDENTITY, “the ‘romantic’ reaction to diversity, giving one’s heart too wholly or readily to another way of life, substituting its norm (real or imagined) for our own”. In today’s linguistics, it is the former failure which is more likely. As an antidote, the typological point of view is able to unveil the importance of EMPATHY, the ‘tool’ for understanding alien cultures and languages. This notion that has so far not played a prominent role (or, in fact, any role at all) in the theory-formation of cognitive linguistics.

In the first decades of the 20th century, Boas and Sapir were wont to contrast the semantic profiles of American Indian languages with those of well-known Indo-European languages. They showed “not only differences in the order and organization of the meaningful elements but also in the matter of what notions are included along with the main content”. For instance, nouns may encode not just number, gender, and case, but also degrees of visibility and closeness; and the encoding of e.g. number, shape, and location may be performed by verbs, not nouns. Only typological linguistics can in this way show how the ‘semantic maps’ of individual languages fill up the ‘conceptual space’ common to all languages (to use terms made current by Croft and Haspelmath, among others). Cognitive-linguistic study of a single language cannot do it.

In the 70's there was a lively discussion concerning the ‘psychological reality’ of generativist (and other) descriptions. It was pointed out at the time that while linguistic descriptions quite naturally tend towards the ideal of economy or axiomaticity, there is no guarantee that this is how the mind of the average speaker-hearer works. In fact, it is rather obvious that it works in a more concrete and redundant way. These topics have been rediscovered in the cognitive linguistics of the late 90's. In this context it needs to be pointed out that cross-linguistic data function as something like a ‘natural experiment’. A recurrent phenomenon carries with it the guarantee of its psychological reality.