I'll explain by an "anecdote" I made up. You're at a party and meet a stranger, who soon seems like an old friend. He agrees with you on politics, sports, religion, books, art, etc. (And déja vu to you, too!) A few days later, at your home, the telephone rings, and a strange voice begins talking. Then the speaker says, "This is Bill. Remember? We met at the party and had a long talk?" IMMEDIATELY YOU RECOGNIZE THE VOICE! Not only that, the next time you speak to Bill on the telephone, you recognize his voice -- even before he says, "It's Bill". Wherefore?

Briefly, the telephonic process is this:

1. The speaker of the telephone has a tranducer which transforms the mechanical vibrations of the human voice into variations in an electrical signal.
2. The electric signal is (analogically) transmitted along the line to the receiver of another telephone.
3. At the receiving end, another transducer transforms electric signal variations into mechanical vibrations that the human ear perceives as sound.

So the receiving person is not hearing a human voice but a simulation of one.

Now, it's easy to show that, even with the improved telephonic technology, an oscillogram of the telephonic signal of "Bill's voice" differs SIGNIFICANTLY from the oscillogram of his actual voice. But you act as if they are "similar". Wherefore?

Three subproblems arise:

1. Why do you think you recognize Bill's voice when you're only hearing a simulation?
2. Why did that "recognition process" begin with that first telephonic cue, "Bill"?
3. Why does that first cue of "Bill" not have to be repeated in the second and succeeding phone calls with Bill?

I'll begin my cognitive-hypothesizing with an "explanation" of "subproblem" #2: WHAT HAPPENS AT THE CUE "THIS IS BILL"?

I say YOU HAVE A MAP (IN THE FORM OF MEMORY) OF THE TONAL INDICATORS OF BILL'S VOICE. The SIMULATED "BILL" PROVIDES THE FIGURE AND THE EVOKED MEMORY PROVIDES THE GROUND. Then THE COMBINED FIGURE&GROUND INVOKES THE REGOGNITION PROCESS. This, then answers "subproblem" #1.

As to "subproblem" #3, the usual "Pavlovian" explanation suffices. Remember? (from ENCARTA) "about 1904. Pavlov found that sounding a bell every time a dog was about to be given food eventually caused a reflex flow of saliva, which later persisted even when no food was produced. Elaborations of this habituative type of reflex are regarded by some physiologists and psychologists as an important basis for many behaviors, both voluntary and involuntary." The signal associated with "Bill" is the "bell", and the F&G of signal-adjoining-memory is the "reflex". (Or jazz it any way you wish!!)

Now, I say that this can be M.E.G. TESTED. As explained elsewhere, magnetoencephalography, or MEG can detect neurological events undetectable by electroencephalography, or EEG, whose signals are obscured by bone and flesh, but these do not significantly obscure MEG signals. The process is so "real-time" that CAN DETECT THE PATH OF A THOUGHT ACROSS THE BRAIN!

By such a tool, all three HYPOTHETICAL STAGES listed above can be "looked for". If one or more of these hypothesized events seems to give rise to an MEG event or track, then this seems to provides some degree of verification of the hypothesized associate.