To appear in: Acta Linguistica Academiae Scientiarum Hungaricae, 1991

like, well, y'know. Garland.
Endnotes

1. An earlier version of this paper was given at the Symposium on Metapragmatic Terms, Budapest, Hungary, July 1990. I would like to thank Robert Harnish and Jacob Mey for very helpful comments on an earlier version of this.

2. I will use the term signal in speaking of pragmatic markers (in contrast to mean in speaking of content material) to mean that the very presence of the form (lexical or structural) com­mits the speak­er to a specific communicative inten­tion. Just as the presence of a lighted green traffic signal signals authoriza­tion for the mo­torist to proceed, and the presence of the bailiff signals the immediate arrival of the judge, so the presence of please before an imperative form signals the speaker commitment to making a request.

3. I leave unaddressed the issue of whether various phonological signals such as sarcastic intonation, emphatic stress, and the like are aspects of sentence meaning or utterance interpretation.

4. Other commentary pragmatic mark­ers signal other types of speak­er comments on the current message, as illus­trat­ed in the following examples.

a) Frankly, we are lost [message has negative import]

b) Reportedly, we are lost [basis for speaker belief]

c) Apparently, we are lost [degree of confidence in belief]

d) Regrettably, we are lost [attitude towards belief]

e) Mark my words, we are lost [attitude towards situation]

These issues are examined in detail in Fraser, 1991a.

5. For the sake of exposition, I am assigning an interpretation to the discourse markers here and below. The points to be made should survive whether or not readers have slightly different readings.

6. In some cases, the discourse to which the marker signals a re­lationship may be in the distant past. For ex­ample, a student in­itiat­ed a conversation with me not long ago with "So, when are you going to Italy?" The so in her utter­ance referenced our con­­versa­tion of some two weeks earlier. For a detailed examination of so. see Fraser 1991b.

7. Commentary markers, other than discourse markers, cannot be absent from the sentence without a loss of meaning. For example, the pres­ence of frank­ly in "Frank­ly, you didn't do very well in the exam" sig­nals a speaker comment, which cannot be in­ferred when frankly is not pre­sent.

8. I see this pro­cess to be anal­ogous to what occurs when one in­terprets good in "a good meal" versus "a good movie" versus "a good boy," or when one interprets just in "just now" versus "just behind the barn" versus "just right." How this process proceeds, however, remains unstated.

9. Aside from the obvious explanations that involve the non-dis­course marker grammatical status of the form (e.g., that and is a coordinate conjunction which occurs primarily in sentence-initial posi­tion), one po­tential explanation for the absence of discourse mark­ers in sentence-medial/final position is the difficulty in dis­­tin­guishing their func­tion from the same formative functioning as a part of the sen­tence content. For example, whereas the po­ten­tial ambiguity of "Now where were we?" can be reduced, if not re­solved, by the presence a comma intonation, this is not possi­ble for the alter­na­tive "Where were we now?", although the dif­ference may be signalled by the utterance-final intonation. I am unaware of re­search which provides an ac­count of these re­stric­tions.

10. The reader is referred to (Barton, 1990) on the issue of el­liptical sentences.

11. There is also the utterance "Because!" in response to "Why aren't you cleaning up your room?" which appears to have become a fixed form, perhaps short­ened from "because I don't want to." In any event, it is not a discourse marker.

12. Some of these pause markers appear to function as a kind of "start-up" form, signalling that the speaker is taking time to think about the answer or at least not responding too quickly, perhaps out of deference to the hearer.

13. The distinction made here are based exclusively on semantic grounds--the pragmatic meaning of the discourse marker, the not on their syntactic privileges of co-occurrence, which do not ap­pear to be useful in the analysis.

14. English focus markers (e.g., emphatic stress, a WH word, the lexical material in the scope of even), which signal the part of the utterance the speaker wishes to make most salient to the hear­er, are different from topic markers and are a type of paral­lel pragmatic marker on this analysis.

15. The lists of discourse markers in the following discussion are intended to be illustrative, not exhaustive. In some cases, a marker belongs in more than one group but has not been included for clarity of exposition.

16. The expression on the one hand is the one exception I have found of a discourse marker which signals that the current mes­sage is related not to a prior one but one forthcoming.

17. For a detailed examination of but, see Bell, 1991.