This goes right back to the grounding of structuralism
in phonological distinctive-feature theory: it is finely dis-
criminated differences between sounds which allow us to dis-
criminate the words of which they form a part, as for instance
bin/pin. But there are great practical difficulties. In phonolo-
gy the units that are discriminated fall within a limited set
of possible phonemes specified by a limited set of phonologi-
cal distinctive features. But if we use this mental process as
a model for the analysis of myth, we need first to agree about
the units of discussion. Within any particular story, how is
the overall text to be cut up into segments for purposes of
comparison? And among several stories which differ in detail
but which seem to have a certain overall similarity, how is
the analyst to decide whether any particular story is or is not
a member of the same set? The structuralists’ failure to for-
mulate any rules about how such questions should be an-
swered seems, on the face of it, to be a very serious defect
in their methodology.
In practice, the procedure is intuitive. For those who
consider that a structuralist methodology is appropriate to
the analysis of religious literature, one obvious set of primary
units consists of all the stories in the Bible, while another
such set is provided by all the stories about S ́iva in the classi-
cal Vedic and Puranic literature of the Hindus. The bulk of
the materials thus specified is enormous; any practical analy-
sis would have to restrict the scope of the investigation still
further by ad hoc criteria. But it is easier to justify arbitrary
limits of this sort than Lévi-Strauss’s universalist practice,
which allows him to put into one set stories drawn at random
from any part of the world and any cultural context.
It should be noted that the arbitrariness of structuralist
analysis connects up with certain of the arguments of
poststructuralist literary criticism. Myths (however defined)
and sacred books and works of art in general, including
music, drama, dance, and the plastic arts, can be regarded
as text without authors. The message in the text is what the
reader (or auditor or participant-observer) discovers. It is
treated as the word of God, but it is an aesthetic response,
something in the reader’s own unconscious mental processes,
which makes the discovery. Thus any “text” is polysemic, a
multiple combination of signs; it has many possible mean-
ings, and no particular possibility has any special authenticity
that the others lack.
This is what makes sectarian diversity in literary reli-
gions so very common. The devotees may all share the same
sacred book, but there are vast numbers of different ways to
put an authentic interpretation on what it contains. Structur-
alist and poststructuralist theorizing have provided a sort of
rational explanation for this all-too-obvious phenomenon of
history.
Prerequisites for structural analysis. A convincing
structuralist analysis of even a very abbreviated set of texts
takes up a great deal of space, so that in an article of this sort
exemplification is hardly possible. The following three pre-
requisites are essential.
- The total text under consideration must contain a
number of separate segments (stories) which differ in detail
but which are also in some respects similar. Taken together,
these similar stories form a set. The items in the set can be
compared and contrasted. A case in point is provided by the
four Gospels of the New Testament. At a certain level each
of the Gospels tells the same story, yet the details differ and
are in some respects radically contrasted. Whereas orthodox
Christian biblical criticism has assumed that these contradic-
tions are of minor significance, or can be satisfactorily ex-
plained away to leave a residual unitary account of what
really happened in real historical time, the structuralist as-
sumes that it is precisely in the differences that the message
of the Gospels, considered as a set, is likely to be found.
Within the total text thus considered, all parts of the text
have equal value; it is quite inappropriate to the method to
discriminate among different kinds of story element under
such labels as parables, historical narratives, and folk tales (see
Leach and Aycock, 1983, chap. 5). - The major segments (stories) must themselves be seg-
mentable into elements (incidents in the stories, motifs).
There is a pattern of relationships between the elements in
any one story. The patterns differ in the different stories. The
analysis calls for a comparison of these differences. In other
words, as in phonology, the coding of the total system is as-
sumed to be a structure of relationships between relation-
ships. - The establishment of the patterns and the contrasts
between them call for close attention to very fine details in
the texts under consideration. Ideally, the analysis should
take account of every detail; it is a presupposition of the dis-
tinctive-feature thesis that, while the text may contain redun-
dancies, it cannot contain accidents. Every detail adds some-
thing to the cogency of the message.
Examples of structuralist analysis. Multifaceted struc-
turalist analyses often turn out to be more interesting than
those confined to a single dimension. As we have seen, the
structuralist thesis is that the fundamental pattern of the
structure under examination is “in the mind.” The patterns
that can be observed out there in the world of cultural experi-
ence—in speech or written text or musical performance or
ritual sequences, in the design of works of art and buildings,
or in the layout of cultural space—are all transformations of
the same mental structure. Thus, at one remove, they should
also be transformations of each other. If structuralist analysis
has application to religious studies, its principal value might
be to give unexpected insight into how the aesthetic imagina-
tion is able to carry out these transformations from one artis-
tic medium to another.All this is very laborious. The linguistic analogy is with
the parsing of a sentence, first into words related in a gram-
matical structure, and then into phonemes related in a pho-
nological structure, and then into the patterns of distinctive
features that constitute the phonemes. The skeptic needs to8754 STRUCTURALISM [FIRST EDITION]