658 DE AUGMENTIS SCIENTIARUM 



Ad Ciphras igitur veniendum. 1 Earum genera baud pauca 

 sunt : Ciphrae simplices ; Ciphrae non-significantibus characteri- 

 bus intermixtas ; Cipbra3 duplices literas uno cbaractere com- 

 plexae; Ciphrae Rotae; Ciphrae Clavis; Ciphras Verborum; 

 aliae. Virtutes autem in Ciphris requirendae tres sunt ; ut sint 

 expeditae, non nimis operosae ad scribendum ; ut sint fidae, et 

 nullo modo pateant ad deciphrandum ; addo denique, ut, si fieri 

 possit, suspicione vacent. Si enim epistolae in manus eorum 

 devenient qui in eos qui scribunt, aut ad quos scribuntur, po- 

 testatem habeant, tametsi Ciphra ipsa fida sit et deciphratu 



orthography does preserve it) up to the present time. For the future, pronunciation 

 would still be free to change, and orthography would still follow ; but the changes of 

 pronunciation would be less rapid and capricious, and the corresponding changes of 

 orthography would be not gradual but immediate. Pronunciation would change, not 

 according to fashion or accident, but according to the laws of nature; and each change 

 would be registered as it came in the printed records of the language. All this would 

 surely be a great advantage, whether we regard language as a medium of communica- 

 tion, for which it serves best when it is most uniform and constant, or as a record of 

 the progress of human thought, for which it serves the better in proportion as capri- 

 cious and accidental changes are excluded arid natural changes marked and regis- 

 tered. 



Bacon was probably thinking of some particular scheme proposed in his own day, 

 in which the existing alphabet was to be used. Many such partial schemes of ortho- 

 graphical reform have been attempted from time to time, all of which may be justly 

 condemned as " useless subtilties," not because the thing aimed at ut scilicet scriptio 

 pronunciationi consona sit would be useless if accomplished, but because, without 

 such a reconstruction of the alphabet as should enable us to assign to each distinct 

 sound a distinct character, the thing cannot be accomplished. With an alphabet of 

 only twenty-six letters, it is impossible to make the spelling of English represent the 

 pronunciation, because there are more than twenty-six distinct sounds used in Eng- 

 lish speech. It has recently been shown, however, that with an alphabet of only forty 

 letters, every sound used in speaking good English may be represented accurately 

 enough for all practical purposes ; and a few more would probably include all the 

 sounds of all the classical languages in Europe. 



Two or three alphabets of this kind have been suggested within the last hundred 

 years. There was one proposed by Benjamin Franklin, another by Dr. William 

 Young, another by Sir John Herschell. But the first serious attempt to bring such an 

 alphabet into general use, and fairly to meet and overcome all the practical as well as 

 all the theoretical difficulties, was made by Mr. Alexander Ellis and Mr. Isaac Pitman 

 in 1848. And there can be no doubt that by means of their alphabet every English 

 word now in use may be so written that the spelling shall contain a sufficient direction 

 for the pronunciation. Nor is there any reason to apprehend that it would ever be 

 necessary to remodel it, since, however the fashion of pronunciation may change, it is 

 not likely that any new elementary sounds will be developed ; and therefore, though 

 we might have to spell some of our words differently, we should still be able to spell 

 them out of the same alphabet. 



As for the fear that, if such a reformation were adopted, works composed previously 

 would become unintelligible, it has been ascertained by many experiments that chil- 

 dren who have learned to read books printed phonetically in the new alphabet easily 

 teach themselves to read books printed in the ordinary way ; and therefore, even if the 

 new system should become universal for all new books, no one would have any diffi- 

 culty in mastering the old ones. J. ] 



1 See, for an account of these ciphers, the appendix at the end of the volume. 

 Bacon's biliteral cipher (see infra, p. 659.) seems, as I have there pointed out, to be 

 connected with one which had been given by Porta, which also depends on the prin- 

 ciple of which the Electric Telegraph is now a familiar illustration, that any number 

 of things may be denoted by combinations of two signs, as in the binary scale of 

 numeration. 



