THE AUTOBIOGRAPHY OF LEONARDO PISANO
RICHARD E.GRIMM University of California, Davis, California
F o r the mathematical h i s t o r i a n interested in biographical details s Leonardo Pisano, better known to r e a d e r s of this journal as Fibonacci, was a frustratingly modest genius. In his extant writings he tells us next to nothing of himself. In only one place, the second p a r a graph of the 1228 edition of his revised Liber Abbaci (Book of Calculation), first published in 1202., does he convey to us information about his e a r l i e r life; and even then the information, given m e r e l y as an incidental backdrop for his explanation of his purpose in writing the Liber, i s very scanty and lamentably lacking in the precision which he displays in his mathematical elucidations. This second paragraph had, in the 1202 edition, been placed at the very beginning of the book; but in the revised, second edition of 1228, Leonardo wrote a dedication to the celebrated court astrologer of Frederick II, Michael Scott, who had requested a copy of the work, and thus this dedication became the work 1 s first paragraph, with the "autobiographical paragraph" following immediately after it. Today 1 s mathematicians a r e familiar with Although only this second, revised edition, since it is the one which Baldassare Boncompagni printed as Volume 1 of his two-volume Scritti di Leonardo Pisano (Rome, 1857-1862). Boncompagni knew of the existence of six manuscripts containing this autobiography, he based his edition — the first, and still the only complete printed edition which we possess — on only one manuscript, the handsome but frequently badly faded Conventi Soppressi C. I. 2616, dated to the early fourteenth century. manuscript. His failure to collate his manuscripts and his reliance upon a manuscript often difficult correctly to read led Boncompagni into an astonishing number of e r r o r s , both o f t r a n s c r i p tion and of punctuation. The brief autobiographical second paragraph is unfortunately not immune from either type of e r r o r ; yet this section forms the basis for most of the statements about Leonardo 1 s early life which a r e found in current histories of mathematics, encyclopedi a s , and special a r t i c l e s . Unfortunately, there has also been a considerable amount of e m broidering upon Leonardo's spare Latinity by many of those who have employed Boncompagni 1 s text — which is to say all scholars who during the past eleven decades have written on L e o n a r d o ' s life. It i s not my intention h e r e to refute point-by-point the many extravagant s t a t e ments found about Leonardo in this m o r e than century-old l i t e r a t u r e . Instead, I wish to p r e sent the second paragraph anew, basing my text on a collation of the six manuscripts which contain it. Following the text I shall offer a translation, along with some footnotes, keyed I hope only that some misconceptions about 99 both to the Latin text and to the translation. Let me state at once that not all the problems in this paragraph a r e hereby forever resolved. This manuscript is now housed in the Biblioteca Nazionale Centrale in Florence; for convenience, I shall hereafter refer to it as Boncompagni's
in quindecim capitulis distinctam componere laboravi. he took charge. siliciam. and further. absque ilia minime inveniatur. while on business 7 . et provinintroductus. to the nine digits of the Hindus. as well as the a r t of P y t h a -
goras I considered as almost a mistake in respect to the method of the Hindus. as a consequence of marvelous instruction in
the art . summam huius libri. quam intelligibilius potui. ut extra perfecto pre ceteris modo hanc scientiam 1 0 appetentes instruantur. m e in pueritia mea ad se venire faciens. mihi deprecor indulgeatur. If I have perchance omitted anything more or l e s s proper or n e c e s s a r y . with its pre-eminent method 10 .
But all this even.eciam. T h e r e fore. ibi me studio abbaci per aliquot dies 2 s t a r e voluit et doceri. Almost everything which I have introduced I have displayed with exact proof. with their varying methods. Vbi ex mirabili magisterio in arte 3 per novem figuras indorum gr. Greece. while adding certain things from my own understanding and inserting also certain things from the niceties of Euclid 1 s geometric a r t . and Provence. et gens latina 1 1 de cetero. Quare. ciam cum suis variis modis. syriam.
Si quid forte minus aut plus iusto vel
nee e s s a r i o interim si. ex proprio sensu quedam addens et quedem etiam ex subtilitatibus euclidis geometrice a r t i s apponens.
Fibonacci can be laid aside and that we can more accurately a s s e s s what his Latinity does allow us to a s s e r t . might be instructed. I have striven to compose this book in its entirety as understandably as I could. following my introduction. 1 2
. multum studium et disputationis didici conflictum 8 . and for it 4 I realized that all its aspects 5 were studied in Egypt. dividing it into fifteen chapters. 1 2 After my father's appointment by his homeland as state official 1 in the customs house of Bugia for the Pisan merchants who thronged to it.100
THE AUTOBIOGRAPHY OF LEONARDO PISANO
[Feb. and. the knowledge of the a r t very much appealed to me before all o t h e r s . Sicily. Syria. fere omnia que inserui c e r t a probatione ostendens. had me in my boyhood come to him and there wanted me to devote myself to and be instructed in the study of calculation for some days 2 . in view of its future usefulness and convenience. in o r d e r that those further seeking this knowledge. since there i s no one who is blameless and utterly provident in all things. Cum genitor meus a patria publicus scriba 1 in duana bugee pro pisanis m e r c a toribus ad earn confLuentibus constitutus p r e e s s e t . ad que loca negotiationis causa 7 postea 6 peragravi p e r Sed hoc totum etiam. and the algorism. et a l g o r i s mum atque a r t e m pictagore 9 quasi e r r o r e m computavi respectu modi indorum. cum nemo sit qui vitio careat et in
T h e r e . and taking s t r i c t e r pains in its study. I pursued my study in depth and learned the give-andtake of disputation 8 . I beg indulgence. inspecta utilitate et commoditate futura. amplectens strictius ipsum modum indorum et attentius studems in eo. sicut hactenus. as they have been up to now. in o r d e r that the Latin 11 people might not be discovered to be without it. scientia a r t i s in tan turn mihi pre ceteris placuit. embracing more stringently that method of the Hindus. omnibus undique sit circumspectus. et intellexi ad illam 4 quod quicquid studebatur ex ea 5 apud egyptum. and at these places thereafter 6 .
the a r t of its exposition. Its vagueness matches the vagueness of the Latin. 4. ilia must refer to suspect i s corrupt. " How much time he actually spent at Bugia in his study Finally.
either scientia. and negotiationis is genitive with loca: " . but ad ill am as a shorthand way of saying ad illam cognoscendam or discendam ("for learning it") is very h a r s h . I feel. The phrase per aliquot dies. though all the manuscripts have it. 5. Did his grecia em-
pueritia mea. " etc. This rendering. which I strongly As it stands. tarn must modify postea. in the latter p a r t of which Leonardo was born. gives us no indication of the amount of time which elapsed between Leonardo 1 s boyhood experiences in Bugia and his travels around the Mediterranean. Just who gave Leonardo this "marvelous instruction" is not stated. but it would be generous to consider F u r t h e r . when we know that he also spent time in Constantinople. Boncompagni 1 s manuscript reads ad queloca negotiationis tarn postea peragraui p e r multum studium et disputationis didici confLictum. The text as it stands offers no basis for much of the standard lore found in biographies of Leonardo regarding his father as ' ' s e c r e t a r y . and the loss of the gerundive early in the manuscript tradition is a strong probability. With this reading. " " m e r c h a n t / ' "agent. is vague indeed. the knowledge of the Hindu system. We simply do not know the precise nature of the position held by Leonardo 1 s father e He was appointed (constitutus)
by P i s a to this post. " "warehouse h e a d . it should be noted that Leonardo uses the word abbacus for "calculation. ." "head of a factory. this was the period of time Leonardo 1 s father wanted him to study the Habacus. the older meaning of abacus as a calculation board h a d grown to include the operations which the abacus performed. brace the Byzantine capital? 6. as he says) when he came to Bugia. to which places of business
. The word for " t h e r e a f t e r . it to imply more than a fortnight. " postea. compels us to refer _ea and ilia to the same thing. My translation is the best I have been able to do with ad ill a m . The difficult quicquid studebatur ex ea.1973]
THE AUTOBIOGRAPHY OF LEONARDO PISANO
1. "while on business." "business man. It is somewhat m y s t i fying that Leonardo mentions these particular five regions as containing all aspects of the Hindu l o r e . a word which we perhaps translate too easily as customshouse." as the present translation r e n d e r s it. or to a r s ." and hence "all its aspects were studied. It i s very probable that he returned to P i s a and went abroad again several It should not be forgotten that he was still a lad (in y e a r s l a t e r . the phrase can be tortured into sense by taking "whatever was studied of it" to mean "all there was of it was studied. but there is no hint of this in the text. It has been frequently assumed that his instructor was Moorish. 2. 7. 3. . This unsatisfactory translation is the most that should be advanced for publicus scriba. namely calculation in general. " By the twelfth century. after reaching maturity. which looks like a r e n dering of the Italian per qualche giorno. Note that Leonardo says specifically that his father wanted him to be instructed for some days in the study of calculation. which certainly involved duties at Bugia (present-day Bugie in Algeria) in connection with the Pis an duana. coming immediately after the strange ad illam. Leonardo does not tell us." i s based on an examination of the six autobiographical manuscripts.
Leonardo. also cryptic. even"). Leonardo utilized the opportunities which his business t r i p s provided to investigate the Hindu number system more thoroughly. etiam should probably be taken with hoc totum (= "all (2) The words etiam et a r e in five of the autobiographical manuscripts but are t h i s . and et algorismum should mean "and the algorismus. " refer to the disputationis conflictum at the end of the preceding sentence? Or does it have as appositive algorismum two words l a t e r ? I doubt the l a t t e r alternative." etc. 603-613). " 8. but also the method of expounding it in scholarly debate. there is deep . Almost certainly.
Tarn postea i s bad Latin for tan to postea. to judge by the There are variety of readings which the manuscripts exhibit at this point. algorismum. and my translation must rely in p a r t on emendation. especially since all the other manuscripts give causa instead of tarn. I think.p o s s i b l y i n c u r able — textual corruption. L e o n a r d o s succinct way of saying " L a t e r . it may be surmised. causa. This is an extremely forced rendering. a l g o r i s m u s . strange. from sed to pictagore.102
THE AUTOBIOGRAPHY OF LEONARDO PISANO
[Feb. r £ . i s a m a r e ' s nest of difficulty which has not been adequately investigated by those who have read it. One is practically forced back to the preceding disputationis conflictum. 9. The phrase i s . pp. mine)." Once again. the great ninth-century Arab mathematician. Had he seen some e a r l i e r work of the Hindus in his travels which made the Arab adaptation seem inferior? Kurt Vogel in his article on Fibonacci in the Dictionary of Scientific Biography (Vol. The Latin h e r e . Peragravi per multum studium I have rendered as "I pursued my study in depth. " The phrase possibly means that in Egypt. Syria. are easily confused. Would Leonardo say that he regarded algorism as quasi e r r o r e m when compared to the methods of the Hindus? (I propose a tentative answer in the next note. l e a n only believe that Leonardo intends some contrast between the Hindu system as transcribed through the Arabs and the "original" system developed in its pure form by the Hindus. Hoc to turn. In the ligature employed by the scribes copying Leonardo 1 s manuscripts in the twelfth to the fourteenth century. ad que loca negotiationis causa post a while on business at these places. the hoc should refer to something under discussion. which Leonardo would then be contrasting with the theoretical basis of the system of Hindu numerals. "all t h i s . If the reading is c o r r e c t . This is a poor contrast at best. (1) Does hoc to turn. algorismum. sought out local scholars on his business t r i p s and mastered not only the theoretical material of the Hindu number system. Leonardo has not previously discussed hoc totum. I
cannot believe Leonardo wrote it. a n d c a .) Again. " would almost certainly be a reference to al-Khwarizmi. speculates. "all this. would Leonardo regard this algorism as "almost a mistake" when compared with the Hindu s y s t e m ? If the text i s kept as i s . the method of argumentation itself.
so much l a t e r I wandered. seems a reference to the medieval practice of discussion and debate on set topics. whose very name was corrupted to "algorism" and referred to the practice of calculating with Hindu numbers. The final phrase et disputationis didici conflictum. tarn. and the other lands he has just mentioned. through [= in the course of?] considerable study. IV. and I am not happy with it. three principal a r e a s of difficulty.
even if Leonardo (and presumably. a noun is needed between atque ("and also") and In the four "noun-less" manuscripts. but this seems unlikely. written in ligature d r c ^ by the scribe. My reasons for so believing and my j u s -
tification for the proposed emendation a r e as follows. I find it difficult to believe that some early scribe would have c a r e l e s s l y omitted a relatively uncommon word like arcus. When algorismus is mentioned by itself. One other manuscript. has already shown himself to be guilty of confusing £ and t_ when he read C& as tarn instead of causa.r£e ? he read drcn9 E. It should be noted that two of these a r e roughly contemporary with Boncompagni 1 s manuscript and that the latter has no special claim to paleographic superiority.1973]
THE AUTOBIOGRAPHY OF LEONARDO PISANO
p. though in extensive checking elsewhere I have found the long us ending for fourth declension nouns such as arcus and gradus written out by the scribe. He might. one (obviously guilty of a slip) adque pictagore. What. moreover.
. Of the six autobiographical manuscripts. the Biblioteca Laurenziana No. written at least a century later than Boncompagni 1 s.
grounds appear to belong to a common tradition. only Boncompagni f s clearly reads a r c u s . script. Hence it is possible that. which on other Could this word have been a r c u s ? In manupictagore ("of Pythagoras"). it would have for Leonardo 1 s r e a d e r s but one reference. The literature on Pythagoras. D. discussed above. three have atque pictagore. reads which could stand for a r c u s . The scribe of Boncompagni f s manuscript. B. (3) The final p h r a s e . they have asked. reconing with lines. 605. To balance algorismum. 783. i s "a scribal concoction. atque a r t e m pictagore. A reference to the " a r c s of Pythagoras" is too esotoric and restricted. without qualifying adjective." makes excellent sense It also s e r v e s as a satisfactory
in context. a r e Pythagoras 1 a r c s ?
compagni f s text r e a d s atque a r c u s pictagore. "the a r t of Pythagoras. A. Certainly artem pictagore. The other four manuscripts all omit the word arcus. since both words a r e common (the word a r s appears thrice in this p a r a graph) and in manuscript m o r e closely resemble each other than do atque a r c u s . so far as I have ascertained. hence a r c u s . and since Leonardo h e r e considers
tant enough to be classified alongside the algorismus. contains no a l pictagore i m p o r lusion to any such p h r a s e . a phrase which has considerably e x e r answer. his readers) knew s o m e thing about Pythagoras which we today do not. and that is to the Hindu system of calculation. it i s logical to assume that he is making a reference to some large category of Pythagorean mathematics which parallels the algorismus. balancing as it does the e a r l i e r mention of the a r s of the Hindus and the immediately following mention of the a r s of Euclid. C. have been guilty of haplography if he had found dty3 3fife\ atque a r t e m . however. it seems obvious that for some reason the word after atque dropped out early. I suspect. the two words would have appeared as dt<?j dnc° . that Fibonacci might mean the l a t e r algorismus linealis. finding something like d. is the last of the three things which Leonardo regards as "almost a mistake" when compared to the Hindu system. cised the ingenuity of scholars.
J r . Hoggatt. 74-81. No. normally sent out to subscribers in November or December. should be given a decent burial. Vol. E. pp. " Fibonacci Quarterly. Oct. since both logically and paleographically it
[Continued from page 90. To m e . i s r a t h e r murky. Wiley. and the manuscripts admit considerable variation. W. though — I freely admit — by no means certain reading. Hoggatt. " Fibonacci Quarterly. Leonardo 1 s name for the Italians.104
THE AUTOBIOGRAPHY OF LEONARDO PISANO
Feb. Vol. a r t e m instead of the arcus of Boncompagni's text. "Diagonal Sums of Generalized Pascal T r i a n g l e s . as a Arcus. Gould. No. this last sentence might well serve as a motto for scholars who write books. please notify Brother Alfred Brousseau St. Vol. ] A PRIMER FOR THE FIBONACCI NUMBERS REFERENCES 1. 2. three of the autobiographical manuscripts have Boncompagni ! s reading. pp. V. 11. 4. 12. . and I have kept it. The Latin h e r e .3. 1968. Marjorie Bicknell. 4. " Fibonacci Quarterly. John Riordan. "Column Generating Functions in Recurrence T r i a n g l e s . 2. 7. No. 6. 5. Leonardo T s humility g r a c e s his genius. 1968. Paul Nil son. "A P r i m e r for the Fibonacci Numbers — P a r t VIII: Sequences of Sums from P a s c a l ' s Triangle. If you have a change of a d d r e s s . 5. V. 221-234. pp. from ut through scientiam. California
. 9. August 1972. If the interpretation of the word which I have given above i s accepted. "Generating Functions for Products of Powers of Fibonacci N u m b e r s ."Fibonacci Quarterly.
however. "A New Slant on P a s c a l ' s T r i a n g l e . 1 (Feb. 10. E. 1971). H. F. I propose. Combinatorial Identities. i s unworthy of serious consideration. Nov. Vol. 1-16.
Renewal notices. However. 6. m o r e reasonable. then. This means that if your address has changed the notice will not be forwarded to you. No. a r e now sent by bulk mail. and Marjorie Bicknell. though other interpretations of the text than the one my translation implies are possible. 1963. 341-358. Mary's College. " San Jose State University M a s t e r ' s T h e s i s . J r . . 3. Mary's College St. 1. Section 4. 1969. pp. April. 1973
balance to algorismum.