though we might be alone as we do so. This talk was given at the 

 American Museum of Natural History, so let's ehoose something 

 from the animal world for the first sentenee. Here it is: NO GNUS 



WERE FOUND IN THE BUSH, BUT SOME WERE SEEN ON THE VELDT. The 



second sentence is more generally familiar because most of us use 

 word processors these days; it is: finished files are the result of 



MONTHS OF SCIENTIFIC STUDY, COMBINED WITH THE WISDOM OF YEARS. 



That is suggestive of hard work, so let's turn our thoughts to va- 

 cation for the third and final sentence. Here it is: the forest ranger 



DID NOT PERMIT US TO ENTER THE STATE PARK WITHOUT A PERMIT. 



We are now going to make things more difficult, by carrying out 

 a small task while reading those first two sentences a second time. 

 We will count up the number of times we encounter a given letter, 

 while again reading aloud, straight through and with no repeats. And 

 we can give ourselves a good start by choosing the initial letter, N, 

 for the first sentence. Here it is again: no gnus were found in the 

 bush, but some were seen on the veldt. We make a mental or 

 written record of the number of Ns we found, and then we proceed 

 to the second sentence, now using the letter F as our target because 

 that is now the initial letter. So here is that second sentence again: 

 finished files are the result of months of scientific study, com- 

 bined with the wisdom of years. As before, we make a mental or 

 written note of the number of target letters — Fs this time — that we 

 counted. 



Our second reading of the third and final sentence is less de- 

 manding; we simply read it aloud once more. Here it is: the forest 



RANGER DID NOT PERMIT US TO ENTER THE STATE PARK WITHOUT A 



permit. As Max Velmans (1991) pointed out, this sentence is inter- 

 esting because it includes two occurrences of the six-letter sequence 

 p-e-r-m-i-t, which we first pronounce permit and later pronounce 

 permit. He suggested that this indicates that human information pro- 

 cessing is not conscious — that it is, on the contrary, unconscious 

 and automatic. I do not agree with that conclusion. If the six-letter 

 sequence had appeared as the first word of the sentence, we could 

 well have found it difficult to know how it should be pronounced. 

 In other words, correct pronunciation requires detection of the rel- 

 evant context. And such detection is not possible unless we have 



