- Facsimiles, or the Duplicity of Duplicates. When talking about facsimiles we need to establish what we mean and also what was the purpose of the operation behind the fabrication of a copy. The most obvious situation is when a mutilated or damaged copy of a book is completed with a facsimile. Most hand made copies are fairly rudimentary and easily recognised, but by the Nineteenth century the art of the bibliographical facsimilist had reached a very high level, so as to be almost indistinguishable from the original. Such almost perfect copies are as dangerous as black ice, especially if we have no warning about their presence, and an essential rule is that any unexplainable phenomenon circumscribed to a single copy of a book should be treated with extreme suspicion. By the standards of the day, however, the work of the facsimilist was not a falsehood but the completion of a copy by another means and the work of an artist such as the elder John Harris (1791-1873) deserves the highest praise; see Barry Gaines, ‘A Forgotten Artist: John Harris and the Rylands Copy of Caxton’s Edition of Malory’, Bulletin of the John Rylands University Library of Manchester, vol. 52 (1969), pp. 115-128; Neil Harris, ‘The Ripoli Decameron, Guglielmo Libri and the “incomparabile” Harris’, in The Italian Book cit., pp. 323-333. A similar artist known as ‘Durer di Padova’, about whom no other biographical information is available, operated in Italy in much the same period, see Neil Harris, ‘The Unicum of the Second Edition of  Boiardo’s Orlando Innamorato and a Forgery of the Last Century’, Rivista di letteratura italiana, vol. 4 (1986), pp. 519-536. Repairs could also be effected by having a local printer set up the missing pages in the type nearest to the original and making a single copy. Although such repairs do not trouble experienced bibliographers and cataloguers, in at least one case I have seen a fellow scholar, who strenuously advocated the photographic reproduction of title-pages in bibliographies, become excited about a variant frontispiece in a a Sixteenth-Century book, until it was pointed out that the ornamental material on the same belonged to the Eighteenth century and that it was certainly a facsimile leaf. But in this particular case the damage was already done. In recent times I have also come face to face with photographic duplicates of missing leaves inserted into exemplars during the course of  a restoration: in such cases the paper is always different and wormholes that suddenly stop when they reach the counterfeit can also be helpful.

A different sort of problem is presented by facsimiles of a whole edition. The practice of reprinting an entire book, usually of the Sixteenth-Century, seems to have been quite common in Italy a couple of centuries later, and of course was not necessarily carried out with an intent to deceive. The best known example is the 1527 edition of the Decamerone reproduced at Venice in 1729, in which the paper of the Eighteenth-Century copy is implausibly thick and robust. In more recent times we have had to cope with photographic reprints of numerous early printed books (I have counted at least six of the Hypnerotomachia Poliphili alone), often done for scholarly reasons, either to make a rare edition more easily available or to reduce wear and tear on a fragile original. One small difficulty I have noticed is that quite often these reprints have a very limited circulation and remain little known, even by specialists, so information about them should always be provided as fully as possible. Photographic duplicates are dangerous when interference takes place with the text, sometimes due to ignorance of the real history of the printing of the original edition: the editio princeps of Castiglione’s Il libro del cortegiano contains in all the exemplars a number of manuscript corrections introduced in the printing shop; in the facsimile edition published in 1986, however, these have been removed, and the same reprint mixes pages coming from two different copies, so that some formes now present a combination of variant states that never existed in the original. I also highly recommend an ironic article by John Flood, displaying an extraordinary mastery of early Twentieth-century printing methods, which explains how retouching to the plates in the 1911 facsimile of Till Eulenspiegel affected the critical text, uncautiously based on the photographic copy rather than on the unicum of the original edition; see ‘“Caveat Lector!” Edward Schröder’s “facsimile” of the 1515 Strasbourg Edition of “Till Eulenspiegel” and the Consequences for Scholarship”, in The German Book cit., pp. 45-59. For an excellent general discussion, with further monitory examples, see G. Thomas Tanselle, ‘Reproductions and Scholarship’, Studies in Bibliography, vol. 42 (1989), pp. 25-54, repr. in Idem, Literature and Artifacts cit., pp. 59-88.

- Fingerprint. In French the empreinte and in Italian the impronta, the fingerprint is a sixteen-character code, divided into four groups, used by cataloguers of early printed books to improve their identification of editions. It was originally conceived in the context of the LOC (London-Oxford-Cambridge) project in the Seventies, but has played a significant role in the Italian census of Sixteenth-Century editions, both in the printed versions of letters A-C, issued between 1985 and 1996, and in the electronic version which came on-line in the Spring of 2000 (http://edit16.iccu.sbn.it). More recently it has been introduced to good effect in the same country’s Servizio Bibliotecario Nazionale (SBN) database for the Libro antico and into the German VD17 project. Trilingual instructions (English, French, Italian) were issued in a printed pamphlet in 1984; see Fingerprints / Empreintes / Impronte, with text in two fascicules, the first containing the ‘Manual’ and the second the ‘Examples’ (Paris, Institut de Recherche et d’Histoire des Textes, 1984), while a version has recently been published in German. The Italian version of the text, with small modifications, can also be found in the cataloguing manual issued for the Census; see Istituto Centrale per il Catalogo Unico, Censimento delle edizioni italiane del XVI secolo: Manuale per la compilazione della scheda (Roma 1987).

For neophytes, the first group is taken from the first recto following the title-page: the first two elements are provided by the last two characters (including punctuation and other signs) in the final line, excluding the signature and the catchword, and the next two are taken from the same point in the penultimate line. The second group is lifted from the fourth following recto, according to the same criterion. The third group is provided by the first recto numbered 13 in Arabic numerals, or 17 if there is no 13; if there are no Arabic numerals, Roman can be used along the same lines; if there is no numbering or no recto numbered 13 or 17 (there are books in which the said numbers fall on a verso), the group is taken from the fourth recto following the second group, again according to the criterion of last line followed by penultimate line. The fourth group comes from the verso of the recto that has provided the third group and, though the characters still come from the last and the penultimate lines, they now come from the beginning and not the end of the line. Particular rules apply to texts in columns [always take the first on the left], to texts surrounded by commentary [as long as it fills the height of the page, still take the first on the left], to very short texts [just keep backing up to the previous group, even if it means taking all the readings from the same page], and so on.

In my opinion the Fingerprint is a straightforward but clever device, which has more than proved its worth in the cataloguing projects I have followed involving early printed books. Once a cataloguer has mastered the basics of the system a Fingerprint can be taken or checked in less than a minute, since over 95% of cases are perfectly standard and require no reference to the manual of instructions, while, despite the simplicity of its construction, it can be surprisingly effective in use. Experienced bibliographers will have no difficulty in imagining both the virtues and the defects of this small glance inside the book. Obviously variants of state or issue, or frisket bite, or poor inking in a corner can affect the Fingerprint adversely, but in most such cases it is easy to analyse and explain the problem. I have also noticed that, even when the Fingerprint has been taken wrongly, with another copy of the same book at hand, it has not been difficult to understand what mistake has been made and to establish that we are not in the presence of a true variant.

The use of the Fingerprint, especially in the Italian Census, has however been attacked or considered of little worth, even by authoritative scholars who should have known better and whose judgments often express an imperfect knowledge of the ways in which early books are made. In particular Enrico Garavelli, ‘Appunti sull’Impronta: catene di edizioni, riproduzioni facsimilari, apografi’, Aevum, vol. 70 (1996), pp. 625-636, has offered a series of sometimes cogent criticisms, brilliantly written yet often irresponsible, above all in their failure to distinguish clearly between cataloguing and bibliographical operations. His article does contain the claim that he merely intends to verify the applicability of the Fingerprint as a tool for textual scholarship, but since the device was never really intended as such it seems a surprising thing to want to do. The sample of editions he chooses moreover, mainly of Italian poetry selected on the basis of the EDIT16, shows the Fingerprint at its least effective, due to the fact that most such editions are line-by-line reprintings, although the prejudice inherent in his analysis is not made clear to the reader.

The Fingerprint was conceived originally as a instument for use in machine-based cataloguing and the availability, from March 2000, of the whole Italian census in electronic form has seen it come into its own. People working in sweatshop early-book cataloguing-operations don’t always have time to sit down and read articles about the utility of the Fingerprint, but few of them will deny that it is an extremely useful working tool. My own experience of library projects has taught me that the greatest difficulties are caused by damaged or imperfect exemplars which nevertheless have to be identified and catalogued as belonging to an edition. Quite apart from the fact that the small libraries in Italy in which I have worked, often with something in the order of a thousand Sixteenth-century books,  rarely have a large assortment of repertories, bibliographies that consent a cataloguer to identify an exemplar that has lost its title-page as belonging to a certain edition are few and far between (to which can be added the idiotic assumption by some scholars that a bibliographical purpose is served by putting in a photograph of the title-page and no information about the body of the book). In the circumstance a search based on the 3rd and 4th groups of the Fingerprint often gives positive results, or, at the very least, a small group of candidates which can subsequently whittled down. Obviously a final check has to be conducted on another copy of the book concerned, but the Fingerprint has already served its purpose.

What use is the Fingerprint to analytical bibliographers? Very little is the most straightforward answer, since it is too simplistic a tool to be useful to the detailed analysis usually applied to the description of printed material. Nonetheless, if I am compiling a preliminary list of editions in some piece of research, I certainly use the Fingerprint as a basic sorting device and I confess that at times it has drawn my attention to points of interest. I believe moreover that it should be included in a bibliographical description as an aid and courtesy to cataloguers, with eventual notes on variant forms and the causes of those variants.

An alternative fingerprinting method, based on the relationship between a chosen signature and letters in the word directly above it was pioneered by David Foxon and has since has been more fully developed in Holland, see P.C.A. Vriesema, ‘The STCN Fingerprint’, Studies in Bibliography, vol. 39 (1986), pp. 93-100. There is no doubt that it is more sensitive than the LOC Fingerprint, though this does not necessarily mean that it is more effective. It has however the disadvantages of a more complex notation, of not having gained widespread acceptance and of not having adapted itself to an electronic environment. In the absence of a large sample of books in which various systems are applied and the results compared, it is not possible to state which is the most effective system.

- Forme (First and Second Impressions) [see Tanselle, ‘Treatment’ cit., pp. 28-31].  It is a self-evident truth that a sheet of paper is printed first on one side and then on the other, or in the language of printers, the white paper and the reteration (or reiteration, also called perfecting). Situations arise in which a bibliographer desires to know which was the first forme to go onto the press, though, in theory, a printing shop can set about the job in several different ways. In consecutive perfecting the whole print run is done in the morning on one side and in the afternoon on the same day (since paper had to be wetted for printing it made sense to make the amount to be done coincide with the working day); in concurrent perfecting two presses are employed and the work to be done is divided into two piles: press one begins with pile A and press two with pile B; when each has completed its pile, press one receives pile B and press two pile A, and they proceed to perfect the work done by the other press. It is also conceivable that, if a very large press run is being executed on a single press, the paper supply may be divided into several lots, each coinciding with a days work: if, on the first day, work begins with forme a in the morning and perfected with forme b in the afternoon, on the second day the shop might just as well continue with forme b in the morning and perfect with forme a in the afternoon. From a bibliographical point of view the outcome would clearly be indistinguishable from a case of concurrent perfecting.

Theoretical models have their uses in moderation, but since the order of the formes through the press is a fundamental part of the history of a book, first forme analysis can provide valuable information for a bibliographer. It has long been applied by incunabulists to establish whether printing in an folio format proceeded seriatim or by two-page formes and thus whether the shop employed a one-pull or a two-pull press: i.e. in the former case the rectos were printed first forme and the versos second forme, in the latter one recto and one verso are first forme and one recto and one verso are second forme. It is also possible that a cancellans leaf might betray itself by having a different order between first forme and second forme.

To detect first or second forme printing, the copy should not have been too severely pressed or hammered by the binder, while if it has been washed and rebound in modern times the task becomes very difficult. If the surface of the page is examined in a raking light, it can be seen where the indentation of the second forme has pushed that of the first forme upwards, though in modern libraries, generally lit with harsh neon from directly above, the operation is almost impossible. It is also possible to construct a simple lamp to obtain the right sort of directional light; see Kenneth Povey, ‘The Optical Identification of First Formes’, Studies in Bibliography, vol. 13 (1960), pp. 189-190. Of course the manuscript rooms in major research libraries are nowadays often equipped with carrels and pencil lights which facilitate the task.

- Forme (Outer and Inner). The order on the press is not necessarily decided by which forme contains the first page of the sheet in seriatim order (since the same forme invariably also contains the last page), though bibliographers talk about the ‘outer’ forme and the ‘inner’ one, and by the first term mean the forme containing at least the two outermost pages of the sheet. For example, in a quarto format gathered in fours, the outer forme is made up by 1r.2v.3r.4v and the inner forme by 1v.2r.3v.4r. For a selection of diagrams of all the principal formats, see Gaskell, New Introduction to Bibliography cit., pp. 88-107.

- Format. Diagrams of typographical layouts are commonplace both in manuals of printing and in those of bibliography, so that most people feel that this concept presents absolutely no problem. But what we normally define as format regards the sheet made at the vat and so, if this differs from the typographical imposition, problems can arise; see G. Thomas Tanselle, ‘The Concept of Format’, Studies in Bibliography, vol. 53 (2000), pp. 67-115.

- Frames, Head-Lines and Running Titles [see Tanselle, ‘Treatment’ cit., pp. 18-24]. The study of head-lines and running titles acquires an almost fetish status in the early years of Studies in Bibliography and is among the distinctive characteristics of the Bowers school in those years. But first let us clarify the relevant terminology. A frame is the box of lines surrounding the text, or sometimes dividing the text into columns, characteristic of the appearance of the page in English (but not only) Seventeenth-Century printing, for instance in the First Folio or in the first editions of Milton’s Paradise Lost. The frame was constructed with brass rules which were easily bent or damaged, so as to acquire a distinctive appearance. The head-line is all the typographical material at the top of the page, including the running title, the number of the leaf or page, and the ornaments. The running title is the text that tells us where we are in the book, so that, if the running title continues from the verso to the opposite recto, we have a single running title but two head-lines.

The importance of these elements for the bibliographer lies in the working habits of the compositor. As has been said, the life of a forme was very brief and therefore, once it had been printed off, it was often taken straight to the press-stone to begin the process of distribution. The first act of the compositor was to knock out and remove the quoins holding the page in place; the next was to lift out carefully the head-lines, which were set momentarily on one side, and then replaced, if necessary with adjustments to the text, in the next forme destined for printing. If, as often happens, some of the components have been damaged or marked in a way that is distinctive, the rotation of the skeleton formes can be established. For instance, if in an octavo format gathered in eights we find the same distinctive running title once per sheet, two skeleton formes are being employed; if it appears twice, once per side, a single skeleton forme is being employed. If a group of running titles disappears for a sheet and then reappears in the following one, the sheet distinguished by their absence may well be a cancellans. Bibliographical history shows that serious study of the Shakespeare First Folio began with the frames and the running titles; see A.W. Pollard, Shakespeare Folios and Quartos: A Study in the Bibliography of Shakespeare’s Plays 1594-1685 (London, Methuen, 1909, p. 134), while, as has already been said, numerous articles in the early years of Bowers’ journal are dedicated to the theme; see in particular Bowers, ‘Notes on Running-Titles as Bibliographical Evidence’, The Library, s. IV, vol. 19 (1938), pp. 315-338. Since then interest has proved more sporadic, but a round up of writings on the subject can be found in Neil Harris, ‘Per una filologia del titolo corrente: il caso dell’Orlando Furioso del 1532’, in Bibliografia testuale o filologia dei testi a stampa? Definizioni metodologiche e prospettive future. Convegno di studi in onore di Conor Fahy, Udine 24-25-26 febbraio 1997, a cura di Neil Harris (Udine, Forum, 1999, pp. 139-204). A word of warning. Identifying running titles, damaged type or watermarks or other elements intrinsic to bibliographical analysis requires time, labour and infinite patience, also because the eye of the scholar has to be trained in an exacting trade. How does a shepherd recognise every single sheep in the flock? By watching them all day long. How does a bibliographer recognise distinctive signs in a book? In much the same way.

- Frisket Bite. A frisket bite is like a love bite. It leaves a mark, it is embarassing and it tells people you’ve been doing something you shouldn’t. Having a frisket or a mask to protect those parts of the sheet that one does not want to come into contact with the furniture of the forme was part of the printing process, certainly from very early days. If however the frisket was badly cut or slipped, it could get between the sheet and the inked type, so that the letters on the edge of the page print in blind. Though never more than a minor nuisance, it can puzzle a cataloguer when the page concerned also happens to be that involved in taking the Fingerprint of the edition. There is no real literature about frisket-bite, but some interesting remarks are made in the debate on the Mainz Catholicon; see Walter J. Partridge, ‘The Type-Setting and Printing of the Mainz Catholicon’, The Book Collector, vol. 35 (1986), pp. 21-52; Paul Needham, ‘Slipped Lines in the Mainz Catholicon: A Second Opinion’, Gutenberg Jahrbuch (1993), pp. 25-29.