This is, of course, an old idea. Huey (1908; reprinted in 1968) devotes two chapters to the role of "inner-speech" in reading. He states: "The simple fact is that the inner saying or hearing of what is read seems to be the core of ordinary reading, the 'thing in itself', so far as there is such a part of such a complex process." (p, 122) While elsewhere Huey states that the fact of inner speech forming a part of silent reading has not been disputed (p. 117), Kolers, in his introduction to the 1968 printing of Huey's book, expresses the kind of ideas that have obscured the relationship between languaging and auding and reading when he states that: "People who read faster than about three or four hundred words per minute, and certainly those who read at rates of a few thousand words per minute, simply have not enough time to form an auditory representation of all they read." (p. xxvii).

Of course, Kolers gives no data to indicate that people can read a "few thousand words per minute." In fact, Taylor (1962) presents eye movement records which clearly indicate qualitative differences between "normal" reading and "reading" at 3000 or more wpm. The latter recordings indicate that the "rapid reading" eyes move in a completely different manner than do the "normal reading" eyes. The latter move systematically to the right across a line of print making three or four stops (fixations), and then make a return sweep to the left margin and begin to move to the right again. The "rapid reading" eyes, on the other hand, may move down the left margin for 10 or so lines, then jump to the right margin for 10 lines or so, then back to the left, and so on, quite clearly doing something other than "normal reading."

Thus, while "skimming" or "scanning" can most certainly be accomplished with printed displays, there is little evidence that readers can, or typically do, read at rates far above the rates at which they can aud or speak (see Edfeldt, 1960; Sokolov, 1972, pp. 202-211, for further discussion and research on inner speech and reading; Carver, 1971a for discussion of "speed reading").

The upshot of this analysis is that much of silent reading appears to involve the conversion of printed symbols into the same type of signing systems used in receiving and expressing oral symbols, which are then converted into, or directly give rise to, conceptualizations. Thus, the representation of meaning directly by written language does not appear to be a typical happening, as some have argued is the case with skilled readers (Goodman, 1973; Smith, 1971, pp. 44-45-again we see here the claim that ". . . trained readers can cover [but not read one by one] many thousands of words in a minute" with no evidence given, and with a failure to carefully distinguish reading from skimming or scanning).