In response to the query by Arash Golzari posted as 
http://linguistlist.org/issues/18/18-1463.html#2 
José-Luis Guijarro wrote: 
 
''[T]he real way we proceed is (1) We try to make sense of whatever 
stimulus we notice, (2) We decode the meaning of the sentence that may be 
the gist of this stimulus, (3) We accept it if it seems relevant for us. 
In other words: We start with pragmatic inferencing from premises offered 
by (1) our chosen context AND (2) from those offered by the decoded meaning 
of the sentence uttered, checking its relevance and if it's ok, then we 
don't look for more. If not, we try again.'' 
 
I understand this as saying that relevance is something on our pragmatic 
menu we need to check, and that we do so quite late in the process of 
interpretation. This, however, seems to be radically different from Wilson 
and Sperber's (1986, 1995) own formulation of relevance: For W&S, ''every 
act of ostensive communication communicates a *presumption* (emphasis mine) 
of relevance'', which means that we don't ''follow'' the principle of 
relevance (as it would apply without exception), and that the principle 
doesn't serve as a final checkpoint but as a first step in one's 
interpretation of the intended speaker meaning: taken for granted that the 
speaker is sane when they communicate ostensively, we assume that what the 
speaker says is relevant, and then, and only then, we make an attempt to 
assign an interpretation to what we hear. As such, relevance causally 
precedes the hearer's interpretation of the speaker's utterance. It is more 
than a mere test tube for one's interpretation of meaning. 
 
Regards, 
 
Ahmad R. Lotfi 
Azad University at Esfahan