the verbchase. More generally, some of the content of encyclopedic information for one item may well be dictionary
information for another.
The proble mis that it is hard and perhaps i mpossibleto draw the linebetween what is in the dictionary and what is in
the encyclopedia for any particular ite m(Bolinger 1965a; Jackendoff 1983; Lakoff 1987; Langacker 1987). For
example, the difference betweenmurderandassassinateis that the latter implies a political motive on the part of the
agent. Is this dictionary or encyclopedia information? If the former, then something as complex as a political motive
can bein thedictionary, and we havehardlysimplified therepertoireof“dictionary semantics.”Ifthelatter, thereis no
distinction between the dictionary definitions of these two words. But this latter conclusion is impossible: linguistic
semantics, if it is to explicate the matching between form and meaning,mustbe able to distinguish the meanings of
these two words.
Jerrold Katz (1980) proposes that the dictionary contain only factors that lead to sharp (or analytic) judgments, and
that any sort of graded (or defeasible) factors belong in the encyclopedia. This too is unsatisfying, since it says that
color distinctions, which are well known to be graded, fall into the encyclopedia. Consequently, on this analysis,Green
things are not coloredis anomalous on the basis of dictionary information (since green is analytically a color), butGreen
things are blueis anomalous on the basis of encyclopedia information. For my taste, this is quite a curious outcome of
the theory. Again, becausegreenandblueare not synonymous, a theory that relegates their semantic difference to
“encyclopedic,”hence“non-linguistic,”meaning, cannot be a satisfactory theory of linguistic semantics.
Here is a more complex case, arising from work of James Pustejovsky (1995) (see section 11.9). How does the f-mind
encode what one typically does with objects, for example that one typically reads books, cooks on a stove, and stores
things in sheds? Beingtypicalactions, they certainly don't fall under Katz's notion of dictionary. My impression is that
most people proposing the dictionary–encyclopedia distinction would consider the mthe beginning of“drowning in
complexity,”hence encyclopedia.
Now, thinkofverbslikefinishorenjoy, whichare understoodas pertainingtoan activity. When these verbs take physical
objects as their complements, as infinish/enjoy the book,finish/enjoy the beer, one readily interprets the min ter ms of these
typicalactionsoneperforms onthem: onefinishes orenjoysreadingthebookbutdrinkingthebeer. Thus, byhypothesis,
a full understanding of these sentences requires encyclopedic information about the objects in question.^141 This
286 SEMANTIC AND CONCEPTUAL FOUNDATIONS
(^141) Finish, but notenjoy, requires that this typical activity have a sequential structure. Thusenjoy the stoveisfine butfinish the stove is odd withoutsome contextual support, e.g.
the understanding that one is cleaning the kitchen.