Re: Binary Oppositions / Fanning

From: Daniel Ria–o (
Date: Fri Apr 03 1998 - 10:53:28 EST

>On Page 57, Fanning makes a point (not original to him) about binary
>oppositions. He states that a binary opposition (aspectual or otherwise) is
>made up of a marked and and unmarked member. No big deal. But then he says
>that the unmarked member is *not* the opposite of the marked member. The
>marked member of the pair has some additional property "A", so we could call
>it "plus A". Then unmarked member of the pair is not "minus A." The property
>"A" for the unmarked member is left unspecified. It is not indicated one way
>or the other.
        Fanning isn't very clear here, but he doesn't' say that what is
made up of a marked and unmarked member", is every binary opposition but
only the *privative opposition". It seems to me that Fanning doesn't always
feel at home with issues of theoretical structuralism. For a very clear
exposition of such issues, see the book of Ruip'erez quoted on Fanning's

>If this analysis is correct, the implications are fairly significant. What I
>am wondering is what kind of evidence can be put forth to support the theory
>that an "unmarked" element is neutral while a "marked" element has a specific
>additional property indicated by its marking.

        Semantic oppositions give clear evidence: take the generic word for
"dress" in Greek: *e)sqh/s*. It forms a privative pair with another term
for a special kind of dress, the *xitw/n*, designating "the woollen shirt
worn next to the body" (the definition of the "baby" Liddell). Clearly
*xitw/n* is here the marked pair of the privative opposition: its
definition contains the note "woollen", while a *e)sqh/s* can be, or can be
not, made of wool.
        The *i(ma/tion*, considered as an "outer garment" stands also as
the marked term of a privative opposition with *e)sqh/s*, but it is just
one member (together with *xitw/n* and many other terms designating
specific kinds of *e)sqh=tes*) of the equipollent oppositions that conforms
the semantic field of "dress" in Greek.
        If life where so simple and binomically perfect, God would be just
the first and most excellent software designer and programmer, but things
in reality are always more complicated: But of course, we must simplify if
we want to operate with things!
        To indicate just one further complication, there is the coexistence
of several "systems", not perfectly integrated, in the same comunity of
speakers at a given time. In the most usual system (let's call it S1)
*i(ma/tion* is the "outer garment". But, on the other hand, it derives
directly from the most generic term for "dress" in Homeric Greek, *ei(=ma*.
Such origins make of *i(ma/tion* a suitable term for any kind of dress, so
that there can be (there actually is, in fact) a S2 where *i(ma/tion*
stands as the neutral term for "dress" in privative apposition with
*xitw/n*, and the "square" (I don't know if that's the correct term for
what we call in Spanish "casilla") for the "outer garment" is filled by
another term, or it remains empty.
        Beside, note that for some theorists the assertion that "the
"unmarked" element is neutral while a "marked" element etc." is not a
theory, but a postulate, and therefore it needs no proof.
        The phonetic systems offer very good examples of the different
kinds of oppositions. They provide the paradigmatic example for other
fields of linguistic, since there the terms have a formal difference. But
oppositions in grammatical systems (like tense, aspect, case, etc.) are
much, much more complicated (I should know: I'm writing a thesis on that).

Daniel Rian~o Rufilanchas
c. Santa Engracia 52, 7 dcha.

This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:39:20 EDT