The Language Log says:
I suspect that people who are happy with "will have had gone" have the reduced form of have (often written "of") in that construction. But there are several obvious questions about the grammar of the rest of the string, which I leave open for discussion in the comments.
For example: if will_of, would_of etc. are finite forms, how come they're followed by had? This might be like "might could", but examples like this are rare:
My guess is that if indeed these speakers have reanalyzed
will have as a single modal
will_of (henceforth referred to as
will_have) that is lacking in its perfective aspect, the situation becomes fairly easy to explain.
There are two posited changes in this Variant English:
1. that past participles may occur consecutively, as evidenced in the original phrase
will have had gone
2. as mentioned,
will have has been reanalyzed as a single modal
will_have, which is a lexical variant of
will and no longer indicative of perfective aspect
As for change #1:
The premise of this analysis rests on the assumption that the sample phrase
will have had gone represents a variation in colloquial speech, rather than a simple error. In Standard English the word
had in
will have had is a past participle, and as there is no evidence whatever to suggest that
had has been reanalyzed as simple past in the Variant English (which would require an additional violation of Standard English rules), we will therefore accept the explanation that violates the fewest rules and maintain that it remains a past participle.
As for change #2:
The premise of this analysis rests on the assumption that
will_have is now a variant of
will as stated above. If we assume that
will and
will_have coexist in the Variant English, and that they serve different functions, however slight such a difference may be, the question thus necessarily falls to what differentiates
will from
will_have, i.e. which situations call for one and which call for the other.
As we all know, in Standard English, modals such as
will cannot be followed by anything other than the bare infinitive, e.g.
will go,
will have gone, and the auxiliary
have requires a past participle, e.g.
have done,
will have gone. There is no evidence whatever to suggest that the Variant English's single modal
will_have has lost the requirement, despite being lexically identical to
will.
Therefore, because,
A. perfective aspect requires the use of
have as an auxiliary,
B.
have can no longer follow
will without forming
will_have and losing its marking for perfective aspect,
and as mentioned,
C.
will_have must be followed by a past participle,
B. past participles can follow other past participles,
Then if we assume that the Variant English has successfully found a strategy to indicate future perfect that works under these restrictions, we can conclude that the only recourse is to employ
will_have before perfect forms (which employ
have as per rule A) in order to protect
have from assimilation into
will as per rule B, necessarily changing it into
had as per rule C. The result is
will have had gone.
