Responding to a review may seem unnecessary and unpleasant, and has often seemed so to me. But circumstantial criticism of specific matters is one thing, specious remarks quite another – especially if such remarks, stemming from a personal and biased reading, call into question the author’s whole philological method, and on occasion come close to insult. May I therefore call upon Ovid’s words, turpe quidem contendere erat, sed cedere visum / turpius.
Hunink is certainly correct in the most emphatic of his substantial claims, that there are typos in the index locorum. I suspect that an attentive reading will reveal an even greater number of inaccuracies, which is unfortunate and a regrettable oversight on my part. (I do take some consolation from my experience as a reader, since I got used to finding this kind of error in those long series of numbers that usually form the indexes: I presume it is sort of a natural characteristic of such treacherous appendices …). I hope this problem will be sorted out in the next printing of the book.1
However, I simply fail to see why this error, that in the end is so common and widespread, should be considered a real crimen to be laid against this book. Unless, of course, one accepts Hunink’s premise, that readers will only consult it rather than read it from cover to cover: on the contrary, I think that a selective reading based on index entries is not an appropriate way to approach the book (even less so by a reviewer I should add). This approach inevitably results in some misunderstandings. Indexes are extremely useful tools but, if I may use a metaphor, they are also ‘blind’. This is perhaps even more true of the index locorum, which puts together disjointed passages, sampled from different chapters of a book and only connected by their proximity in the text.
Were this work a lexicographical one, Hunink would be right in considering the numerical typos in its index a gross negligence rather than a venial sin. However, this book is neither a lexicon nor a catalogue. It is a new study of Apuleius’ style which aims to offer an original key to the interpretation of his work as a whole. Examples are used as supporting evidence for my argument: they are not the argument. I have started from a hypothesis, tested my idea through a close scrutiny of the Latin text, and finally reached a conclusion about what I believe to be a crucial aspect of Apuleius’ writing. A reader who wants to evaluate this book fairly is actually expected to read it from cover to cover. Then again, the book can certainly be consulted as a repertoire of remarks on individual passages, and this option is offered and made easier by the presence of indexes. Yet, it remains a monograph on an extremely important aspect of Apuleius’ linguistic experimentalism and is by no means a collection of stylistic oddities. It is any reviewer’s right, of course, to agree or disagree with an author; but it is not fair to describe the book under review as something different than what it is.
I feel bound to clarify at least the major points where misunderstandings seem to arise from such a reading.
The first problem with relying on the index locorum as a guide through the book is that one finds listed there all kinds of word play (etymological puns, as well as phonic word plays, puns of great effect, and those already noticed by commentators), not to mention all the passages that do not actually include any word play. It is therefore ungenerous, to say the least, to repeatedly point out that “even a beginning reader [sic] of the Latin” would have picked up some of them. In many cases, moreover, the point is not the original discovery of a new word play. For example, it may well be true that even a dilettante would spot the enallage at Met. 11,5,2 fluctuantes Cyprii. This is not a momentous discovery, I agree. Nevertheless, a more extensive reading of the relevant chapter would have revealed the significance of my argument: that the frequent occurrence of enallages like this one, which cause a semantic shift of adjectives (sometimes such a strong shift that it challenges comprehension), should be considered before adopting emendations of such passages as met. 5,23,6 detectae fidei, 6,28,5 compta diligentia, and 8,7,7 adfixo servitio. In all these cases the text is evidently beyond the interpretative abilities of an amateurish reader of Latin, so much so that several talented philologists have tried to emend it. Listing a series of instances serves therefore a purpose in its own right, independent of the significance of every individual passage quoted.
A second consequence of such a ‘desultory’ reading style is that one might expect to find an instance of word play where there is none, and feel disappointed as a result. This is exactly what happens to Hunink in the case of Met. 11,1,1 (analyzed on p. 55 of my book), where he rightly remarks that “there is no pun involved”. No pun is involved, indeed, nor do I see why there should be one. The passage at met. 11,1,1 is brought into the argument for a different reason: it serves only as a parallel to another passage, met. 1,2,1, where the same verb emergo occurs. In this case, the reviewer has another and more serious objection to my method: he clearly does not approve my use of the adjective ‘normal’ (inverted commas by Hunink) with reference to the usage of emergo in 11,1,1, to support the necessity of an emendation at 1,2,1, where the transmitted text is unproblematic for Hunink (I quote: “The ‘normal’ use of emergo with separative ablative is adduced … as evidence against F’s reading, emersi me, transitive with accusative”). I cannot understand how one can consider it “outrageous” or simply methodologically wrong that an author’s usus scribendi (which, incidentally, is perfectly consistent with classical standards) is used as an argument against a text that appears grammatically unsound. The passage at 11,1,1 ( emergentem… fluctibus) is only one of the several passages used as evidence for the consistent Apuleian usage of intransitive emergo, construed either absolutely or with the ablative of origin. The phrase emersi me looks simply impossible to me, and so it has to such distinguished predecessors as Leo, Helm, Robertson, and more recently Keulen, who have emended the transmitted text in different ways. I only add that the simple expunction of me does not solve the problem: notwithstanding the validity of Keulen’s arguments (see his comm. ad loc.), I have a serious problem with the transitive usage of emersi, governing the previous accusatives ardua montium et lubrica vallium et roscida cespitum et glebosa camporum. I respect of course position of Hunink, who sees no problem in this. I merely pointed out that Vallette’s elegant conjecture, «emensus» emersi, which provides a good solution to the syntactical peculiarity of the phrase, is also supported by Apuleius’ propensity to this type of paronomasia, with which I specifically deal in this section of the book (the paleographic explanation of the error is self-evident).
Last but not least, a third consequence of such a casual approach to the book is a gross (and, I have to say, definitely insulting) misunderstanding of my philological method. Hunink repeatedly states that I support more or less recent conjectures against the transmitted reading “because they fit a specific category of puns”; he declares himself “disturbed” by this. He implies that my textual choices are whimsical and arising from a stubborn wish to demonstrate my hypothesis. Perhaps it is true that I do not stand in much awe of the Codex Laurentianus (F), but I do have a great respect for philology, its methods and its principles.
For the benefit of readers who are less familiar with the textual tradition of Apuleius’ works it may be useful to recall that F is, in all likelihood, the progenitor of all the available Apuleian manuscripts, and that F was written more or less nine centuries after Apuleius’ time. Is it really absurd to question the readings of F, not only when they are clearly corrupt, but also when the rules and conventions of the Latin language are apparently above suspicion? There must be, of course, good grounds for emendation: whenever I support or suggest an emendation, I am always motivated by some inconsistency (be it linguistic or logical) in the transmitted text.
This is particularly evident in the case of the passage pointed out by Hunink. The reading of F in 11,30,4 deserviebat is certainly corrupt: an enormous number of emendations have been put forward to restore the passage, and among them I support one ( ibidem serebat) that implies a word play which is also attested elsewhere in the novel. This perfectly reasonable proposal by Oudendorp, already suggested by Beroaldus, is far from being hazardous; indeed, it is clearly better than many other conjectures, as a glance at Helm’s or Robertson’s apparatus will confirm.
But there is more in the book that suggests that the text of F should perhaps be challenged where it has never been – and this is precisely what displeases Hunink, who protests: “it struck me that the author puts perfectly acceptable manuscript readings into question”. This is absolutely true. And I hope this can be regarded as one of the original aspects of my book: it is certainly one that I am proud of.
I shall mention only a couple of cases in which the text of F is not usually called into question: the conjectures conserentes for the transmitted conferentes at Met. 5,15,3, and polentarium for polentacium at Met. 6,19,2 (pp. 128-129, 137-141 in my book) should at least be mentioned in the critical apparatus. I have never emended F to adapt Apuleius’ text to my reading and to an aprioristic classification of puns. On the contrary, I devised my classification a posteriori, in order to catalogue the different kinds of etymological puns (real variations on a theme) that Apuleius uses so frequently. This peculiar stylistic feature, Apuleius’s penchant for etymology, is the object of my book: I hope I succeeded in showing the continuous presence of it and how it works in the text. This peculiar trait of style can be used – such was my original aim – like any other trait of Apuleius’ usus scribendi : it can help us restore and interpret difficult parts of the text, and can guide us towards a correct evaluation of scholarly contributions. This is my method, which I very much regret escaped Professor Hunink.
In my opinion, much can still be done to improve the text of the Metamorphoses and of the other Apuleian works; much is to be expected of M. Zimmerman’s forthcoming Oxford edition. I also think that much can be done on the grounds of our always improving knowledge of Apuleius’ language and style. We should not let ourselves be caught in a dilemma between a prudently conservative and a boldly innovative stance. Whenever the text of our codex unicus is suspect for reasons of grammar and language, or when doubts exist about its meaning, respect for the paradosis should not prevent us from supporting a good conjecture, if it is based on Apuleius’ idiolect and style, finds good parallels, and is paleographically easy to explain.2
Unless one wants to consider the codex Laurentianus 68,2 a unique exception among classical texts, or to think that the philologist’s trade is completely pointless.
1. Typos in the indexes are admittedly bothersome, but surely there are worse mistakes. One example: the misprinted Latin text humani generi (sospitatrix) (p. 143 in the book – not far from one of the numeric typos mentioned by Hunink) is a linguistic monstrum that escaped not only my attention, and the editor’s, but also that of the reviewer, who quotes the phrase exactly as he finds it on the page.
2. It is a really democratic method, if one thinks about it. The very same consideration of the usus scribendi can in fact conversely support the readings of F against unnecessary conjectures, or contribute to the correct interpretation of uncertain passages. Several instances of both cases can be found in my Index rerum.