306   Linda /?. Waugh
status of the * + * or *-* is arbitrary — it is rather the [signalization of x] - [nonsignalization of x) that is telling for the internal structure of the opposition.)
Second, many investigators have assumed that because the unmarked may have the zero-interpretation, exactly in those contexts where the unmarked and marked are not (explicitly) opposed, then the zero-interpretation must reflect a •neutralization*, a suspension of the opposition of the two terms (on the model of phonology). However, this is a confusion ofphonology with semantics. Because the phonological distinctive features have no other meaning than "differentiatedness', in those cases where substitution (differentiation) is not possible, the opposition is 'neutralized' and is no longer valid. In grammatical and lexical semantics, however, the meaning of a given unit is based on the conceptual categories associated with it and is thus not dependent on substitution in all contexts. Indeed, one would expect that in certain contexts substitution would not be possible, because there would be a 'semantic congruence' between the given item and the context in which it is found. Communicators do generally want to convey coherent messages. So cases of nonsubstitution are in fact to be taken as strong evidence for the meaning of a particular item, not as evidence for the lack of meaning..
Further linguistic examples
Other examples of marked ~ unmarked relations in grammatical meaning include the following:7 (1) The relation of number for the noun in the English grammatical system is defined by the opposition of'more than one'~'one', where 'more than one' (the plural) is marked. Since the singular is unmarked, it may be used in both the minus-interpretation (one book, a man came—'not more than one*, i.e.. 'one') and the zero-interpretation (man is an animal, the elephant is a strange beast, is someone there?, who's at the door? = 'nonsignalization of more than one'; number is for the most part nonpertinent or unknown: cf. Jakobson 1957). (2) In many languages the personal pronouns are related as third person (unmarked) versus both first and second person (with a marked reference to the 'participants of the speech event'); in addition, in the opposition of second to first person, the first person is marked since it specifically refers to the 'producer of the given message' (cf. Benveniste 1946 and 1956, and Jakobson 1957). The unmarkedness of the third person may be seen in such examples as he who hesitates is lost, would the gentleman like to order wine? — usages showing that the third person does not invariably exclude a reference to the speech situation. The unmarkedness of the second
Marked and unmarked   307
t
|   person (as against the first) is shown by examples such as the 'general you':
you never know, to go from Morrill Hall to Rothschild's you turn left at
í   stewart Avenue: and speaking to oneself: you dummy. This pronominal
*   example also shows that a 'three-way* system is often characterized by the
i  intersection of two different (binary) oppositions. (3) In French, for a
I   large part of the nominal system, the masculine gender is unmarked and
■   the feminine gender is marked. Evidence for the masculine being the
I   unmarked gender in French is the fact that it can stand for the category as
\   a whole: e.g., lionne 'lioness* specifies the female, while lion lion' may
,   include both female and male. In addition, if there is an adjective
:   modifying two nouns, one masculine and one feminine, the adjective is put
,;   in the masculine plural: des hommes et des femmes intelligents 'intelligent
;   men and women*. The masculine is also the gender used when the gender
\  is unknown: ily a des lions lá-has "there are some lions over there'; or, e.g.,
j   in questions : qui est mart? 'who has died?'; or when the gender is not at
;  issue (e.g., when adjectives, verbs, and adverbs are made into nouns in
'•   French, without any other reference, they are masculine: le beau 'the
!   beautiful", le parier "(way of) speaking', le 'vraimenť que tu viens d'utiliser
j  'the "truly" that you've just used'), in fact, a phrase like le beau 'the
beautiful' refers either to a previously mentioned noun in the masculine
singular (minus-interpretation) or to a male being who is beau (minus-
|  interpretation) or to the simple nominalization of the adjective beau (zero-
I   interpretation). (4) In Russian, the relation of perfective ~ imperfective
:   aspect is defined by the opposition 'completion (accomplishment) of
f  action'~'action in itself. In indicative mood usage, the perfective is the
marked term (focus on completion of action) and the imperfective then
either conveys the noncompletion of the action (minus-interpretation) or
I says nothing about completion (zero-interpretation). In imperative mood
usage, there is a 'markedness reversal*: the imperfective is the marked term
•  (focus on the action itself) and the perfective then either conveys the
j completion of the action (minus-interpretation) or simply does not focus
!  on the action (zero-interpretation) (see Karcevski 1927; Jakobson 1932,
1   1957). This example shows that the mark itself is independent of the
1 opposition, in the sense that either pole of the opposition may take on the
mark, depending on the context in which the opposition is used. Marks
are not absolutes, but rather are constantly defined in terms of nonmarks
I and in terms of the context of which they are a part.
,     Marked - unmarked relations in lexical meaning include the following:
'  (1) In the lexical opposition homme versus femme in French, homme is
j unmarked not only because of its gender, but also lexically — just as in
i English (see above), it may be used to stand for the category as a whole
\ [i'homme eel inconnu — man the unknown). (2) In the opposition of the
I