One of these was Gerard of Cremona, not of course the Cremonese who died in 1187, but the younger scholar of the same name, perhaps a son or nephew of the elder. He is distinguished as Gherardus de Sabloneta Cremonensis. The Victorine manuscript[162] supplies evidence that he contributed to the work in which Michael Scot was now engaged.
It is not impossible that Philip of Tripoli may have joined in the new enterprise. His name does not indeed appear in any of the manuscripts which contain the Latin Averroës, but we have seen that he was certainly in Spain about this time and even at work with Gerard of Cremona.[163] His intimate relation to Michael Scot is also beyond question, and, upon the whole, it seems reasonable to suppose that the Emperor may have engaged him to help in the work now going forward.
However this may have been as regards the exact details of time and persons, we may regard it as a matter now for the first time brought to light and established, that in the years between 1217 and 1223 there existed a college of translators in Toledo just such as that which had done so much excellent work there a century before. In the new school Frederick II. held the honourable place of patron, as Archbishop Raymon had done in his day, while Michael Scot and Gerard of Cremona aided each other in completing the version of Averroës as Dominicus Gundisalvus had lent his help to form that of Avicenna. This view of the matter should be found very interesting, not only in itself, but with regard to the conclusions arrived at by Jourdain, whose discoveries in the literary history of the twelfth century it so remarkably repeats and extends to the following age.
This correspondence between the earlier and later schools of Toledo is even more close and exact than we have yet observed. It appears also in the fact that a Jewish interpreter was attached to each, and rendered important service as a member of the college. Under Don Raymon this place was held by Johannes Avendeath, or Johannes Hispalensis as he is commonly called, who worked along with the Archdeacon. ‘You have then,’ says Avendeath, addressing the Archbishop, ‘the book which has been translated from the Arabic according to your commands: I reading it word by word into the vernacular (Spanish), and Dominic the Archdeacon rendering my words one by one into Latin.’[164] The same division of labour seems to have been followed in the new school which Frederick promoted. The Emperor drew the attention of these learned men to Averroës, and signified his desire that a version of this author should be prepared like that which had been made from Avicenna. Michael Scot and Gerard of Cremona were responsible, the former probably in a special sense, both for the general conduct of the undertaking, and, in particular, for the accuracy of the Latin. Now these scholars also, like their predecessors, availed themselves of the help of a Jewish interpreter. This was one Andrew Alphagirus, who seems to have taken the same part that Avendeath had formerly done, by translating the Arabic of Averroës into current Spanish, which Scot and his coadjutor then rendered into Latin.
Such at least appear to be the suggestions which offer themselves naturally to one who peruses the colophon to the copy of the De Animalibus ad Caesarem preserved in the Bibliotheca Angelica of Rome. Thus it runs: ‘Here endeth the book of Aristotle concerning animals, according to the abbreviation of Michael Scot Alphagirus.’ The form of expression is curious, but may be exactly matched from the versions produced by the earlier Toledan translators: that is, if we are to believe Bartolocci. This author, in the first volume of his Bibliotheca Rabbinica, mentions a manuscript of the Fondo Urbinate in the Vatican which, he says, contains the four books of Avicenna on Physics translated by ‘Johannes Gundisalvi.’ This name has evidently, like that of ‘Scoti Alphagiri,’ been formed by composition from those of the two translators, Johannes Avendeath and Dominicus Gundisalvi who aided each other in the work.[165]
As to the personality of Alphagirus, the only ground of conjecture seems to be that supplied by Romanus de Higuera, who, speaking of the learned men assembled in 1218 at Toledo for the astronomical congress, mentions that one of them was ‘el Conhesso Alfaquir’ of Toledo.[166] The place, the date, and the similarity of name, are all in favour of our supposing these two to be one and the same person. Nay further, as Alfaquir was of Toledo, and did not need to be summoned thither in 1218, there is no reason why he should not, as the ‘Alphagirus’ of 1209, have assisted Michael Scot in producing the De Animalibus for Frederick.
It is from a remark made by Roger Bacon that we know the first name of the Toledan interpreter to have been Andrew, and that he was a Jew. Bacon gives us this information in no kindly spirit, but in order to lead up to the bitter conclusion that Scot’s work was not original, but borrowed from one whose labours and just fame he had appropriated. ‘Michael Scot,’ he says, ‘was ignorant of languages and science alike. Almost all that has appeared in his name was taken from a certain Jew called Andrew.’[167]
A sufficient answer to this serious accusation may be found in what we already know of the literary fashions of the day, and, in particular, of the traditional methods of work pursued by the Toledan translators. It was precisely thus that the Archdeacon Gundisalvus had used the aid of Avendeath. A little later too, we find the same system adopted in the translation of the Koran promoted by Peter the Venerable. That ecclesiastic thus expresses himself in sending a copy of his book to St. Bernard: ‘I had it translated by one skilled in both tongues; Master Peter of Toledo; but since he was not as much at home in the Latin, and did not know it as well as the Arabic, I appointed one to help him … Brother Peter our Notary.’ To his Koran Peter the Venerable joined a Summa Brevis of the Christian controversy with the Mohammedans. This work also came from the pen of Master Peter, and with regard to it he makes the following remarks: ‘By giving elegance and order to what had been rudely and confusedly stated by him (i.e. by Master Peter) he (i.e. Brother Peter the Notary) has completed an epistle, or rather a short treatise, which, as I believe, will be very useful to many.’[168]
This correspondence throws a clear light upon the case of Michael Scot in regard to the charge of plagiarism. Like Master Peter, he was familiar with both the Latin and the Arabic language. His weak point, however, we may suppose to have made itself felt with regard to the latter, which he probably knew better in its colloquial than its literary form, and this must have been the reason why he availed himself of the aid of a Spanish Jew to secure the accuracy of his work. Such collaboration seems to have produced nearly all the previous versions which came from Toledo, and it is obvious that the honour due to the various contributors who combined in forming these translations can only be determined by those who have it in their power to make a careful and unprejudiced valuation of their individual labours in each case. We may gravely doubt whether this was what Bacon did before he sat down to pen his sharp censure on Michael Scot. Certainly such an estimate is now out of the question. We can only affirm the undoubted fact that the critic was wrong when he said Scot did not know Arabic. The contrary appears, not only from the probability we have already drawn from his Sicilian residence, but by actual testimony of a very honourable kind.[169] Nor must we forget to notice that the openness with which this copartnery was carried on affords a proof that no deceit could have been thought of in the matter. Considering the past history of the Toledan School, it must have been taken for granted that every version which came from thence under the name of a Christian scholar owed something to the care of his Moorish scribe.
Even had we not been able to make such an appeal to the use and wont of the times in vindication of Scot’s method of work, might not a little consideration of what was natural and inevitable in such a task have served to explain what Bacon found so objectionable? The scholars from distant lands who came to Toledo could not, as a rule, afford to spend much time there, and were anxious to use every moment of their stay to the best advantage. They naturally therefore secured on their arrival the services of a Jew or Moor for the purpose of learning Arabic. Needing a knowledge of that tongue not so much in its colloquial as its literary dialect, they must have been engaged from the first in the study of a text rather than in conversing with their teachers. What then could have been more suitable than that these scholars should begin by attacking the very books of which they desired to furnish a Latin version? This method had the merit of gaining two objects at once. The students learned to read Arabic, following the text as it was translated to them by the interpreter. Writing in Latin from his vernacular, and polishing as they wrote, they engaged from the day of their arrival in the very work of translation which had brought them to Spain. It is plain too that any modification of this method which the case of Michael Scot might demand would depend on the knowledge of Arabic he already possessed. It must therefore have been such as left him more and not less credit in the result of his labours than that which commonly belonged to the Christian translators in Toledo.