348 JOURNAL OF LAW AND POLICY
recognizability would be possible at the syntactic level. The
issue is to find, again borrowing from phonetics, some invariant
signal among the variation and noise (in an information theoretic
sense).^46 Or borrowing from statistical terminology, what
syntactic patterns would be distinctive enough among the
potential note writers to differentiate intrawriter variation from
interwriter variation?
Language is a conventional behavior where for the sake of
mutual understanding we share the same code. In information
theoretic terms, each of us is both sender and receiver. This is
how we manage to finish each other’s sentences: we are using
the same code we share with another person in our linguistic
circle. So the notion that individual language is unique, or that
each of us has a unique linguistic behavior, is an idea that
linguistics as a discipline denies by the very definition of
language as a conventional behavior and shared code.
Even though linguistic behavior cannot be literally unique, it
can and does show variation. By definition, dialect is the name
for group-level linguistic behavior, where subgroups within the
language can be determined. At the individual level, linguistics
has posited the notion of idiolect, or a variation of language at
the individual level.^47 Clearly, idiolect cannot be a unique
language, or, again, the unique language would have a speaker
of one, but variations at the individual level might still be
discoverable. Idiolect was first posited at the phonetics level.
The biological substrate of phonetic articulation certainly makes
phonetic individual differences feasible.^48 Idiolect later became a
useful theoretical term in recognizing syntactic variation between
syntacticians. There is still no empirical method for
demonstrating that each person has his or her own idiolectal
variation that is uniquely identifiable, but author identification
merely has to recognize intrawriter vs. interwriter variation
(^46) Cf. CLAUDE E. SHANNON & WARREN WEAVER, THE MATHEMATICAL
THEORY OF COMMUNICATION (1971).
(^47) See, e.g., FROMKIN ET AL., supra note 32.
(^48) See id.