No Images

M. Clarke, The Archaeology of the Book: Formulating Analytical Research Questions, e-conservation Journal 2, 2014, pp. 10-16
Available online 28 May 2014
doi: 10.18236/econs2.201403

The Archaeology of the Book:
Formulating Analytical Research Questions

By Mark Clarke


In addition to being vehicles for the transmission of texts and images, books are also archaeological artefacts, and an archaeological approach benefits both conservators and historians of the book. Chemical analysis can, for example, confirm provenance, distinguish hands within a manuscript, and identify alterations due to aging or deliberate modification, in the latter case by identifying anachronistic elements.

This is a golden age for the study of the physical properties of works of art in general and manuscripts in particular. Increasingly sensitive analytical instruments have become available, and there has been a renaissance in the philology of artists’ recipe books. This has resulted in a greatly improved understanding of workshop methods and materials. Nevertheless, considering the very large amount that has been written about the history of books, their manufacture and their decoration, there have been proportionally very few analyses published. The study of the materials used for making books is in its infancy compared to e.g. those for easel paintings. To establish what was conventional usage for different periods,  regions, or ateliers, much more work is needed to build up a substantial and statistically significant corpus of analyses.

Unlike other research fields, in codicology (‘the archaeology of the book’), in conservation, and in ‘technical art history’ there are rarely natural structures in which research topics arise, such as those in academic research groups. More prejudicial still, unlike many fields in physical sciences and humanities, the interdisciplinary nature of the technical examination of books means that when someone does have a concrete question, outside collaboration is almost invariably required. The formulation of an answerable question therefore requires cross-disciplinary ‘interpreters’ who can explain the requirements and possibilities of their own specialisms. The best programmes of analysis have been done by physicists and chemists working closely with conservators, librarians, and art historians.

This paper suggests how achievable research questions may be framed, and how the results of these investigations may be disseminated in such a way as to inform wider questions than the immediate concerns that drove the initial research.

1. Choice of Purpose

Research questions in the analysis of book materials are rarely driven purely by conservation questions, and more rarely still by purely art historical questions. In this, codicological analysis differs from analysis of paintings, where conservation and authenticity are the two principal drivers. In practice, the analysis of book materials has, with a few notable exceptions, come as close to pure research as imaginable in that it has been driven almost entirely by curiosity.

Research questions do divide broadly into two overlapping areas, namely history and conservation.

Historical questions attempt to ascertain when an artefact was made, where, and by which individual or workshop, and to establish which parts of that artefact are original and which parts were added later (restoration, forgery, or in the case of a book simply several booklets bound together for convenience). But the combination of results from many books should contribute to wider knowledge of developments in the history and technology of the book, the organisation of workshops, of trade and of exchange, and resultant sociological changes.

Conservators seek to avoid, retard or reverse  degradation and damage, while at the same time preserving the archaeological material evidence embodied in the object. Conservation questions thus may be divided in two: either concerning individual artefacts (condition and state of preservation, risk assessment for display or digitisation, characterisation of ink likely to degrade, assessment of suitable treatments), or combining the examinations of individual artefacts to answer more generally applicable questions (what can we do about ink corrosion, or pigment flaking, or red rot, or the brittleness of acid paper?).

Questions that address both historical and conservation questions simultaneously ask firstly how a book or decoration or illumination was executed (pricking, underdrawings, paint layers, etc.) — all indicators of the conformity or independence of individuals with traditions and the relation of an artist to his materials — and ask secondly in what way the present appearance reflects the artefact as originally produced (discolouration, losses, repairs, etc.), perhaps with a view to returning the object to a more authentic or pristine state.

2. Choice of Books to Examine

In my opinion it is unfortunate that the majority of analyses of the materials of books has been carried out on exceptional books. While it is understandable that researchers (and funders) will want to work on the remarkable, high-profile, newsworthy items, rather than everyday ones, in consequence we remain largely ignorant about standard manufacture. It is as though the history of architecture was written based entirely on  examinations of Stonehenge, the Leaning Tower of Pisa, and the Eiffel Tower. In contrast, analysts of paintings (technically the closest comparable field to that of book analysis), through years of routine examination and routine analysis of a tremendous variety of material (painted by masters and hacks alike), have built up a considerable knowledge of standard fabrication practices. We are thus in a position to appreciate exceptions to those standard practices, and so to derive useful conclusions about age, provenance, authorship, and degree of post-manufacture alteration. In comparison we remain remarkably ignorant about the materials of the conventional early book. And yet it is precisely an appreciation of what was standard practice that increases appreciation of the exceptional efforts that produced non-standard ‘master’ works. All books, and especially all book decorations and materials, however low quality, deserve attention: indeed it is the simple material that forms the bulk of surviving material and will thus (i) most likely produce the most statistically significant evidence and indicators for dating and provenance, and (ii) present the greatest conservation burden (consider, for example, the nineteenth century acid paper brittleness).

3. Choice of Analytical Techniques

Rather than formulating a research question and then commissioning the appropriate analysis, research questions in the analysis of book materials have often been formulated based firstly on what equipment is available. It seems to be putting the cart before the horse, although excellent results have nevertheless been achieved. There are always logistical problems getting books and instruments in the same room: equipment can often not be moved, and books can be too fragile or valuable to be moved. It is on occasion possible to borrow instruments: for example, the Eu-ARTECH funded project ‘MOLAB’ offers a mobile collection of portable equipment ( But it is to be hoped that curators will come to realise that the benefits of a well-directed programme of instrumental analysis are considerable, and that the risk of moving books to laboratory facilities can be made minimal.

2 web
MOLAB analyse a manuscript at the Fitzwilliam museum.

4. Sampling Versus ‘Non-destructive’ Analysis

Historically, the main obstacle to the study of book materials has been the aspiration not to take samples. In consequence, for many years, the only analysis possible was visual examination. But recent advances in non-sampling instrumentation have permitted far more reliable analysis. However, there can still be significant problems.

1) As previously mentioned, there are often logistical problems bringing books and instruments together.

2) Non-invasive methods cannot analyse all types of materials; notably most organic pigments require sampling.

3) In situ non-sampling analysis seems the ideal from a conservation perspective, but not necessarily: for example, clamping a page in place such that an instrument may focus on it can be more damaging than sampling would have been.

4) Layer structures are important, from mediaeval miniatures to modern coated papers, but non-sampling analysis only detects either the top layer (in reflective or surface techniques, e.g. visible light spectroscopy or Raman spectroscopy) or may combine all the results from several or all of the layers (in penetrative techniques, such as X-ray techniques).

5) Many pigments were used mixed (a fact abundantly clear from historical technical texts), but mixtures can be very difficult or impossible to understand without sampling, especially where inorganic and organic pigments are mixed. Often analysts have detected an inorganic pigment, but failed to detect or even consider possible organic pigments mixed in.

In conclusion, sampling is routine for most types of artefact, remaining rare only for books. But modern analytical techniques can be effective with samples so small as to be only visible under a microscope. The damage-to-knowledge ratio is very favourable, and a well-planned set of samples can be used for years, and sent around to several laboratories, for many instrumental analyses, including those not yet invented.

5. Putting It All Together: Framing Questions

The main research categories are therefore history and conservation, and may be either specific questions about individual books, or general questions about many books.

5.1. Example Questions About Individual Books

To simply identify, for example, all the pigments in a book is no longer sufficient. While such routine analysis of course remains essential and must be done (routine analysis often throws up unexpected and surprising information, notably intrusive and anachronistic elements), more sophisticated questions must be asked of the data so obtained.

Dating. In our present state of knowledge, the simplest and most obvious questions, such as ‘when was this book made?’, are the hardest to answer. The unavoidable imprecision of terminal dates for the introduction and deprecation of materials such as pigments (such dates being best expressed as a range of probabilities dependant on the rate and spread of adoption, and typically embrace several decades), means that material analysis is unlikely to provide a more precise dating than would conventional palaeography.

Intrusive elements. Glaring anachronisms, such as pigments only introduced after the supposed date of a book, can indicate later re-working such as retouching or even a complete forgery.

Localisation. An identification of unusual or rare materials will be helpful in narrower localisation (probably by identification of locally-sourced  organic pigments, or locally preferred species    of animal used for parchment or leather).

Manufacturing processes. It has been possible to separate elements of apparently homogeneously produced books by the analysis of materials.   Examples of such successful analyses have included changes in paper within a twentieth century diary (indicating revisionist elements added later), and the consistent use of two different pigments on the same decorated manuscript showing how the work had been divided between two teams.

Grouping. It is frequently attempted to group individual books to certain common locales of production (workshops or individuals) by means of stylistic and historical evidence. Analysis can of course help. If the materials are the same (the notable materials, that is, not the standard ones) in books that are supposed to originate in the same atelier, this can strengthen the connection, whereas inconsistency in material use will weaken it. Similarly analysis can be used to confirm a common origin of disjecta membra: for example DNA analysis of the membranes has been used to reunite fragments of the Dead Sea Scrolls.

5.2. Example Questions About Many Books

It is a good idea to have a large-scale research question, such that each individual analysis of a book can not only inform us about that book but also contribute to a more global understanding of book history. The ‘master question’ is always: ‘what was used, how, where and when?’ This will lead to a better understanding of workshop practices, and more sociological questions such as the demands and expectations of patrons and consumers, and so forth. But this can only be answered by a considerable programme of coordinated analysis to determine statistically significant trends.

Certain questions need specific programmes of research other than the ‘master schemes’ outlined above. An excellent example is the question ‘Is the degradation in these books due to their manufacture or their storage environment, past and present?’ One outstanding project to address this is the British Library ‘Identical Books Project’, where different libraries’ copies of the same 400 printed books are analysed thoroughly for indicators of degradation: pH, colour, lignin content, sulphate content, bending brittleness, etc. The study of environmental conditions can further be carried out by the use of surrogate book materials, that is, historically accurate reconstructions based on historical technical treatises and craftsmen’s recipe books.

5.3. Interdisciplinary Formulation

Analysts and art historians rarely appreciate each other’s issues. Patience is needed, and a willingness to pay attention. The key to designing a research programme is an iterative consultation process.

The book historian, curator or conservator does not want to know much technical detail, just (a) ‘can this technique answer my question and if not, which technique can you suggest?’, and (b) ‘will it damage my object?’ Decisions on the ratio of risk of damage to useful information can thus be arrived at iteratively.

Analysts, however, are often surprisingly interested in the questions of art history and codicology, both historical questions and questions of the history of book technology, and will usually benefit from a fairly full statement of a problem: they may then be able to suggest an individual technique, a combination of complementary techniques, or a statistical approach. Furthermore, analysts are often interested in finding novel applications for their instruments, and in stretching the limits of what their instruments can do (the conservation preference for non-sampling or micro-sampling being a good example).

5.4. Meta-analysis

Common in medicine, practiced in archaeology, yet rare in conservation, ‘meta-analysis’ studies (or ‘systematic reviews’) critically examine the data of previously executed studies to answer new or larger questions. Such studies consider the validity of the methodology, the specificity and accuracy of analytical techniques used, and the actual results, so as to perform a ‘meta-study’ or ‘virtual experiment’ on a far larger range and number of subjects than would be possible for  an individual team or project to do1. More should be attempted in the field of book studies, to add greater statistical validity to what has largely been an ad hoc series of analyses.

5.5. Timetabling and Paying for Analysis

Without specific project funding, the analytical burden typically falls on the conservation department of a library. It is thus subject to considerable pressure and limitation, but nevertheless the conservation environment is in many ways ideal for analysis: there is a conservator present to guard against hazardous procedures or inappropriate handling, and books and instruments can generally more easily be introduced to each other in a pre-conservation state, especially when the spine is broken (permitting flat opening) or better yet, when a book is pulled and disbound2. Samples can be removed before conservation treatment (to prevent contamination by conservation materials) and saved until suitable analytical opportunities become available.

Analysis can be seen as super-documentation, and the opportunity for analysis (or at least sampling) should be seized at the moment of conservation intervention: it will generally be a once-in-a-life-time opportunity to apply otherwise physically or logistically impossible analytical techniques. How good it would be to see libraries prepared to supplement the budgets conservation departments to incorporate such routine analysis?

6. Suggested Priority Questions

I would like to identify a few areas which I believe could usefully be prioritised.

1) The identification of inks and pigments is still a valid and useful area of research, but both the analysis itself and the questions asked of the results need to be more subtle and more comprehensive: rather than continue to be content with an identification of, for example, ‘verdigris’, we must investigate the composition, corrosive tendencies and relative stability of different forms of verdigris made according to different recipes.

2) The importance of organic pigments should receive more attention. These are the most vulnerable to fading or damage by inappropriate treatment, and the degree of visual alteration is not well understood or quantified. Furthermore, their relatively greater variety means that they have potentially the greatest use for identifying ateliers and identifying places of production.

3) Recent advances in micro-sampling and immunological assays allow the analysis of binding media and adhesives with unprecedented precision. This is interesting for the history of technology (the sometimes doubted addition of ear-wax to glaire, as described in mediaeval treatises, has only recently been experimentally identified), but more so for informing conservation choices.

4) Proper analysis of parchment species is to be encouraged: a great many codicological conclusions have been published that are based on what I am confident will prove to be incorrect identifications (the visual differences are often misinterpreted, and furthermore it is not commonly appreciated that the physiology of mediaeval goats and sheep were far more similar than today). The question of ‘uterine vellum’ versus the use of small animals would be one historic debate it would be well to resolve once and for all, and would provide useful localisation data.

5) The importance and consequences of layered paper and paint structures has been under-researched and is under-appreciated. For example, in both western and Islamic manuscripts it seems that organic glazes (thin layers added on top of paint) were common. We know almost nothing about these, yet such organic glaze layers are highly vulnerable to cleaning, washing, and light-induced degradation.

6) Large-scale digitisation brings new questions: we should investigate what happens to the digitised book that will now be far less frequently consulted: whether pages will become blocked or offset, whether a lack of regular airing will encourage mould growth, and so forth.

7) Finally, I would strongly suggest that far greater attention be given to everyday, standard or low quality books, such that we may build up a better picture of common usage.

7. Applicability and Audience of Results

The applicability and audience of research results varies. A detailed examination of an individual object is of immediate concern to those responsible for its curation and conservation, while the more general questions are naturally more interesting and contribute more to the disciplines of ‘technical art history’ and the ‘archaeology of the book’. One of the greatest challenges facing technical codicology is how to make results of specific examinations of individual books widely available in such a way that results may be combined to allow generalisations to be made. When an analysis has been done, every effort should be made to publish the results, in however abbreviated a form: for example as an appendix to a catalogue, or as an appendix to a yearbook, or on a webpage. This need not be an elaborate text, with an introduction, discussion and conclusions, but a simple, telegraphic even, statement of technique and results, perhaps in a simple table of results per manuscript: the purpose not being a formal academic paper but a simple dissemination of data with which future syntheses may be built.

8. Conclusions

Analysis is most successful when it complements traditional codicology, or when it answers a specific conservation question. The amount of analytical work done has increased recently, as increasingly sensitive and unambiguous techniques have been developed. Nevertheless, much research has been ad hoc, and it is still true that not nearly enough is known about the materials of books. Regardless of any academic interest, we must make such analysis a priority to know what conservation treatments will be safe and suitable. Furthermore, the recent dramatic increase in demand for exhibition of books may destroy in a few weeks pigments that have survived for centuries.

The present relative scarcity of book analysis, perceived and genuine, leads on occasion to negative comment (typically by art historians) as to the worth of analysis. This attitude has two manifestations: ‘if nobody does it, how can it be useful?’ and ‘if there is so little comparative data, what use is the analysis of my few objects of interest?’ Each leads to vicious circle of non-analysis. But the pioneering has been done: what is needed now is to make analysis mainstream and inevitable.

While not every feature of every book will ever be fully analysed, whatever analysis can be done contributes to the overview. Analysis can be seen as super-documentation: ideally a way needs to be devised to build in to routine examinations some forms of routine, rapid protocols of simple and inexpensive analysis.

Mark Clarke
This email address is being protected from spambots. You need JavaScript enabled to view it.