In this paper I present an analysis of the distribution and functions of two Finnish prenominal modifiers, sellainen+NP ‘such NP’ and tällainen/tuollainen+NP ‘this kind of NP/that kind of NP’ in spoken language. By means of a detailed analysis of naturally-occurring discourse, I show that existing analyses of these modifiers, while not incorrect, are insufficient. According to my analysis, sellainen and tällainen/tuollainen act not only as ‘new entity/new concept’ markers as suggested by Vilkuna (1992) and Dasinger (1995), but also as markers of non-prototypicality. They can warn the hearer that the subsequent noun should be interpreted only as an approximate characterization of the intended referent.

In formal Finnish, sellainen (1) is usually used to refer back to already-mentioned entities or states (Dasinger 1995). Tällainen/tuollainen (2) indicate "class membership without requiring exact equivalence between the items located in the class" (Dasinger 1995: 188).

(1) En ole syönyt [sellaista ruokaa]. ‘I have not eaten [such food].’ (Karlsson 1999:139)

(2) [Tällaisella autolla] ei voi ajaa. ‘One cannot drive in a car [like this].’ (Karlsson 1999:139)

In colloquial Finnish, these words also have other functions, which are the focus of my paper. Let us first consider sellainen. According to Vilkuna (1992), ‘sellainen+ NP’ is used in colloquial Finnish when introducing entities or concepts that are unknown or unfamiliar to the listener. Dasinger (1995) adds that speakers use sellainen when referring to previously unmentioned/unidentifiable entities, and also when introducing entities or concepts whose characteristics the listener may not be familiar with. The use of ‘tällainen/tuollainen+ NP’ in spoken discourse has, as far as I know, not been discussed in the literature.

On the basis of a detailed analysis of spoken narratives, I suggest that previous accounts of sellainen and tällainen/tuollainen are insufficient and fail to fully capture the conditions under which these modifiers are used. My claims are based on data from 29 Finnish adults (ages 18-58), who narrated the story of a wordless children’s picture book to a listener who could not see the pictures. Thus, I was able to analyze 29 versions of the same story, which allows for more detailed comparisons speakers than is usually the case. Slightly over half of the narrators (15/29) use the modifiers sellainen and/or tällainen/tuollainen, and the most frequent users tend to be the younger speakers.

The speakers’ use of the modifiers reveals a number of interesting patterns. As Vilkuna and Dasinger suggest, sellainen is indeed used to refer to previously unmentioned entities (3) – but so is tällainen/tuollainen (4). (Note the dialectal variants: semmonen = sellainen; tommonen = tuollainen.)

(3) se … kiipes [semmoselle korkeella kivelle]

‘he … climbed onto [sellainen high rock]’

(4) tää …. peura juoksee [tommosen jyrkänteen] reunalle

‘this …. deer runs to the edge of [tuollainen cliff]’

However, not all new entities are marked with sellainen or tällainen/tuollainen. A comparison of the narratives reveals that, on the one hand, some referents, when introduced for the first time, are never marked with sellainen or tällainen/tuollainen. On the other hand, there is a group of referents that, across speakers, often get marked with sellainen or tällainen/tuollainen. For example, when mentioning for the first time a log that the characters climb over, 11 narrators use either sellainen and/or tällainen/tuollainen.

These patterns raise a number of questions: Why are some (but not all) new referents modified with sellainen or tällainen/tuollainen? And more specifically, why do the modified referents tend to be rather familiar concepts (rock, cliff, pond, log)? According to my analysis, sellainen and tällainen/tuollainen can function as markers of non-prototypicality. They warn the hearer that the image immediately evoked by the noun may not be exactly appropriate, i.e. that the actual referent may not be a prototypical exemplar of the class evoked by the noun. For example, the cliff mentioned in (4) is a grassy area within a forest that suddenly slopes down steeply – whereas the protopical cliff is an overhang of rock, often by the sea. By using tuollainen, the speaker indicates that the subsequent noun jyrkänne ‘cliff’ should be interpreted only as an approximate label for the referent. This ‘approximating’ nature of sellainen and tällainen/tuollainen can be derived from their common suffix –lainen. This suffix, when attached to an NP, creates a new word meaning ‘like NP, similar to NP, resembling NP’(Ikola 1986:91). When the suffix is extracted from sellainen and tällainen/tuollainen, we are left with se ‘it/this,’ tämä ‘this,’ and tuo ‘that’. Thus, compositionally, sellainen and tällainen/tuollainen mean ‘resembling this/that/it.’

In sum, on the basis of my analysis of sellainen and tällainen/tuollainen in spoken Finnish narratives, I argue that these modifiers can function as markers of non-prototypicality, as well as new entity/new concept markers.