164 Douglas J. Glick
terms through which the indexical ones are discovered and vice-a-versa. They are mutually
interdependent. We return to both of these points in closing.
To this point in the analysis it would appear that a relatively fixed and stable discursive
construction has been located. It is one constituted by a set of common propositions and an
overlapping, second-order indexical regularity. In the name of a fuller understanding of the
social life of discursive constructions, however, methodological issues, and the related
theoretical implications that they entail, now force us to problematize this apparent
conclusion. The problem is not that we have failed to exhaust the regular indexical life of this
discursive construction because, at least in part, the indexical side of this (or any) discursive
construction is dialectically the result of our methods for isolating out propositional
regularities. Rather these problems should force us to see how, in general, all analytical
representations of a discursive practice, both ours and that of any other interested parties in
practice or in theory, are relative to a particular perspective. There can be no ̳pure‘
unmediated representation of a discursive construction. This is not to make the extreme, and
ultimately absurd, claim that overlap and thus degrees of shared discursive reality do not
exist. They do. Without them social communication would lack the presuppositional grounds
on which it is based. It is however to claim that analysts create methods that are in principle
no more or less privileged than those used by individuals and organizations with very
different perspectives, methods and interests for representing ̳how people regularly talk about
something‘. In closing this argument then, we will return to the data and work through a
series of related moves that demonstrate this point. First, we will consider propositions that
were vulnerable within the focal discursive construction based on this study‘s methodological
assumptions. Second, we will move outside the results discussed above and consider a few
causal propositions that weren‘t found frequently enough to be considered ̳common‘.
Returning first to a denotational approach, note how two of the common causal
propositions reported (n. 35 and 36 found in the right-hand column in Tables 1 and 2 above)
have yet to be mentioned. Like the others, they too were found to be common and thus
members of this discursive construction. They were, however, relatively more isolated from
reflexive interconnections than the others. That is, their presence was neither modeled on an
implicit or explicit theme or theory within the data nor was it patterned from above by a more
widespread ideology. As such, references to it were typically stated in or as the causal
accounts themselves explicitly. In this sense, while still of course legitimate common
members of this discursive construction, their relative isolation makes them somewhat
vulnerable participants. This vulnerability, seen in semiotic terms, suggests the influence of a
strong indexical component to these propositions. For those individuals sensitive to the sites
from which this data was drawn, they stand out as indexically marked propositional elements
in the regular discursive construction. For example, some propositions are more likely to be
associated with particular kinds of speakers/writers or time periods and as such be interpreted
as reflecting narrow ̳political‘ interests (cf. n. 35 on outdated macho values being associated
with relatively older Israelis). Alternatively (as evidenced in n. 36 on Israelis‘ agitated
psychological state), component propositions can index other discourses that restrict the
reflexive support they find within a focal discursive construction. Mutual constraint and
support, as these cases demonstrate, are not the only indexical relations found among the
component parts of discursive constructions.
For our purposes here, we‘ll focus on the second causal proposition: an ̳agitated
psychological state‘ offered up as an explanation for negative, but nevertheless