options for locating the agency of the shifts in the narrating-I’s voices: one or
more of the successive “Frankies” or the more mature narrating-I who appears
at the beginning of the memoir. The first option is implausible since it would
lead to such hard-to-process conclusions as “three-year old Frankie decides
to yield the floor to five-year old Frankie,” and the second option goes against
Smith and Watson’s interest in freeing the narrating-I from a fixed temporal
position, a position that McCourt seems to share.
In addition, saying that the narrating-I is responsible for the shifts in his
or her own voices also assumes that the narrating-I always performs the dual
functions of telling and of orchestrating the telling. This assumption may be
justified with some self-conscious narrating-I’s, but it effectively rules out uses
of a genuinely naïve narrating-I, since the orchestrator would have to be only
pretending to be naïve. But McCourt and others have shown that memoirists
can not only employ genuinely naïve narrating-I’s but also use them in the
construction of very moving narratives.
If, however, we add the implied author to Smith and Watson’s schema, then
we have a much more persuasive and elegant model of autobiographical narra-
tion, one that allows us to unpack the complex “project of self-narration” and
the temporal connections between narrating- and narrated-I’s. The historical-I
of the real author can remain ultimately inaccessible (though I would suggest
that in many cases we can draw probable conclusions about the historical-I);
the narrating-I retains the potential for maximal dynamism; the narrated-I
retains a similar potential as well as its multiple functions as protagonist; the
ideological-I remains the concept of personhood available to the narrating-I
in her cultural moment; and the implied author becomes the orchestrator of
the shifts in the narrating-I’s voices, in the various representations of the nar-
rated-I, and of their temporal interrelations. This implied author can choose,
within the limits of her own imagination as both constrained and stimulated
by her cultural moment, to employ any kind of narrating-I—one in a fixed
temporal location, one whose temporal location constantly shifts, one who is
naïve, one who is sophisticated and self-conscious, and so on—and can show
the stages of the narrated-I’s temporal progression in any order. What’s not
to like?
From Smith and Watson’s perspective, what’s not to like can be found in
what I detect as the two deeper reasons for their polite rejection of my pro-
posal, reasons shared by many others who are skeptical of the concept of the
implied author: (1) Adding the concept to a model that already includes four
autobiographical I’s seems to violate Ockham’s Razor. It’s this reason, I believe,
that underlies Smith and Watson’s objection to the alleged redundancy of the
concept. (2) Adding the concept opens the model to a significant role for
198 • CHAPTER 1 0