Main Menu  


Reflections -part 2

Reflections on Book Structures - part 1

by Pete Jermann     c. 2004

   Herein  I pass on the reflections of an addled, middle-aged guy who spent 23 years working in an academic library binding and repairing books. If you are looking for thoughts on fine binding, you are looking in the wrong place. If you are looking for thought provoking (that means I may well be wrong) concepts related to book structures than read on.

    My job as preservation officer and bookbinder was to extend the life of the circulating collection through repair and rebinding and to do what I could with the time that actually wasn't left over to keep the rare book collections from deteriorating past their current state. Many of the processes I developed and the thoughts I thought were formed by the pressures of a job where the work always exceeded the resources available. I was also in the unique position where I worked in one of the few libraries, if not only library, in the country at that time (and up through 1999 when I departed) that was still doing much of its own periodical binding in-house.

    I began as a student in 1976 and learned the basics under Father Joe Ruther, a Franciscan Friar, who had operated the bindery since World War II. We had two processes for binding periodicals. If the periodical was in anything remotely resembling signatures they were sewn through the fold on sawn-in cords. If you are wondering what something remotely resembling a signature is then you need to experience a burst binding where much of the back of the signature is burst to allow glue to penetrate to the inner signatures.  Burst signatures leave almost nothing for the sewing thread to hang on to. To my constant frustration Father Joe insisted on treating these as signatures nonetheless. If a periodical was comprised of single pages adhesive bound we would drill through the margin and side-sew the binding.

    Probably 75% of our binding hours were spent breaking books down into either individual signatures or individual pages and then resewing them, either through the fold or side sewn. After sewing your first five to ten thousand books the therapeutic value of sewing begins to wear off. That warm, fuzzy, back-to-your-roots, proletarian feeling gives way to the thought that there has got to be a better way. When Father Joe was forced into extended periods of medical leave I began to play. Upon returning from his first six weeks leave he found a new glue, PVA, and the rudimentary beginnings of a new process, double fan-gluing. Though he never adopted nor explicitly approved of what I did, he did assure me that he once was a young man also and, as such, had his period of experimentation. From this point on there were now two ways of doing things, his way and my way. Within several years he passed on and the bindery became mine alone for the next 15 years.

    When I left the bindery in 1999 to stay home with my newest daughter and to tend to TeMPeR Productions I left behind an operation that did very little sewing and lots of fan-gluing. The transition from where I began in 1976 and what I left in 1999 involved much musing, problem solving and major mental shifts. I worked in one place and with one collection long enough to enjoy my successes and to see my failures return to me. From my experience I present below the following often rambling, possibly overstated,  and definitely arguable reflections. Please forgive me for a binding history that is at times fast and loose. My sense of binding history is based largely on books that have passed through my hands. It is not my intent to discourse on the history of binding but to use it loosely to illustrate and explain binding structures.

Sewn bindings vs. Adhesive bindings

    Sewing signatures and gluing single leaves together both solve the single problem of leaf attachment or how to keep the pages together in a book. It is the beginning of the bookbinding process and a single component of the more complex structure that is the finished book. Their success or failure is usually more dependent on the remaining structure of the book than on the original method of leaf attachment. Neither process can lay claim to an inherent superiority over the other. Both come in many variations and both have their successes and failures. With many modern papers either process can be used successfully and if well done, interchangeably. Depending on the paper, some books are better sewn and some are better glued. Experience best determines where that line falls (sorry, there is no easy answer here).

    With the exception of numerically insignificant no-glue bindings coming out of some conservation shops and comb/spiral bindings, all modern bindings are adhesive bindings whether they are sewn or not. Every commercially sewn binding has been glued up on the spine and lined with some manner of reinforcement, be it paper or cloth. Those who proclaim the superiority of the sewn binding have not been paying attention to the work flow across their benches. Once the adhesive fails in a modern sewn binding the sewing will fail shortly thereafter if use of the book continues. A modern, machine sewn book is a loose affair before the glue consolidates the signatures into a manageable entity. Once the glue fails the sewing becomes a destructive liability as the threads, now free to move, tend to saw the needle holes larger and rip the signatures. Sewn books with slick papers often depend on the tiny bit of glue that has worked its way into the thread holes. Should the glue not penetrate sufficiently, the inner leaves will move, ripping against the entrenched thread. Without the protection of an adhesive, a modern sewn book will not successfully stand.

    Please note that in matters of sewing I distinguish modern from ancient. Many books survive from the middle ages with their adhesive long gone and their sewing intact. These survivors are raised as examples of the superiority of thread over glue. However, what they actually illustrate is that if you build a book with pages of animal based leaves (parchment/vellum), which are virtually untearable, or of heavy handmade papers using a thick thread combined with handsewing that creates a tight textblock, you can create a durable book based on sewn leaf attachment. However, these books were very expensive, comprised of materials largely unavailable today, not particularly user friendly, producible in small quantities only, and, consequently, only available to a very small and elite group. The ability to make books available and affordable to the masses (that's you and me) depends on the technologies and materials that produce the modern book. Our attempt to produce durable bindings must be built on what actually exists as opposed to wishful thinking based on a past that is no more.

    Whereas the virtues of sewn books are often uncritically extolled, the defects of adhesive bound books are uncritically broadcast. A poorly done adhesive binding lends drama to its failure in a way a sewn binding does not. Everybody seems to have remembered at least one adhesive bound book that literally burst apart on its first opening while ignoring the intact phone directory that they use repeatedly without failure. Similar to sewn bindings, adhesive binding refers to a broad class of bindings that vary by the type of glue, the manner it which it is applied and by the remaining structure that makes up the bound book. (Note: pet peeve coming up) The use of the term "perfect binding" to refer to all adhesive bound books improperly services any discussion of adhesive binding by imposing negative connotations on adhesive binding. I believe the term "perfect binding" was originally a trademark referring to the very process that produced those dramatically explosive adhesive bound books. Its use to refer to all adhesive bound books provides no useful information and denigrates many well bound books.

    The truth is that, overall, adhesive binding technology is a major success. After a bad start (i.e. "perfect binding") it has come a long way in reducing costs and putting durable books into the hands of consumers. For every adhesive bound book that fails there is an overwhelming preponderance of those on our shelves that never come to our binderies.

    Adhesive binding offers the binder a tremendous efficiency in binding over sewing. In a small, unmechanized bindery the process of leaf attachment takes minutes. Many books that are sewn can be just as successfully adhesive bound. However, this is not to say all books should be adhesive bound. It has its limitations and its problems. There are good adhesive bindings and bad adhesive bindings. Books with very thin papers, such as bibles are probably better served by an adhesive binding which grips each individual page along its entire length then by a sewn binding where the inner pages of the signatures tend to slip against the threads. Book on heavy coated papers are best served by a sewn binding. Between these two extremes there are many books well served by either option.

On Sewing Structures

    The purpose of sewing signatures through the fold is twofold. The first is to secure the gathered pages of a single signature to itself. I refer to this as the intra-signature sewing. The second purpose is to attach one signature to the next, or inter-signature sewing. The two functions, intra and inter-signature, fill structurally distinct roles. Whereas the failure of intra-signature sewing will almost always lead to a binding failure, i.e. loose leaves, the failure of inter-signature sewing is irrelevant in a modern hybrid binding that is both sewn, reinforced on the spine and glued. Glue and super in a modern binding structurally replace the inter-signature sewing. Only when glue and super fail does the inter-signature sewing even come into play and at that point it is probably a danger to the signatures as discussed above.

    The inter-signature sewing in a modern, machine sewn textblock is actually quite loose, allowing significant play between signatures. Add glue and the textblock tightens considerably as the adhesive secures the many signatures into a unified textblock. The addition of a super reinforces this. Glue and super remove the play in the textblock, preventing the inter-signature sewing from being extended to the point where it plays any structural role. To prove this point to a conservator friend of mine, I once constructed a sewn binding with no inter-signature sewing. I sewed each signature as a single pamphlet, aligned and clamped the collection of signatures, glued up the spine and, when dry proceeded to round it, back it and  bind it as if it were a normal sewn textblock. When I presented her with the finished product without her knowing the details of construction she could not discern any difference between it and a normally sewn binding. Fifteen years later that sample is still intact and functional.

    So, what is the significance of this structural distinction between inter and intra-signature sewing? Understanding this can significantly simplify repair of a sewn binding as well as offer new options for sewn bindings. If a sewn textblock separates between signatures into one or more blocks, it can be repaired by ensuring that the intra-signature sewing is intact, stacking the various parts neatly, clamping them together, re-gluing and re-lining the textblock. Securing the intra-signature sewing can be as simple as having just enough loose thread emerging onto the spine to be caught and secured by the glue. Also new bindings can be built from a collection of signatures internally secured as noted above.

On supers, sewing supports and endsheets

    The super, also known as the crash or mull, reinforces the spine and secures the cover to the textblock. Early binders integrated sturdy cords or thongs made of various animal skins into the sewing of the spine. The cords and thongs, apparent on early bindings as raised bumps or bands on the spine of the finished book, extended beyond the spine and were laced or woven into the boards to secure them to the textblock. What we see as the finished binding was then built upon this structural backbone. The tightback construction of early bindings, like the super in modern bindings, further reinforced the board attachment.

    In the nineteenth century demand for books increased. Filling this demand required faster and cheaper methods. As machine sewing had yet to dawn, binderies streamlined the handsewing process and turned to mass production based on cheap labor. Binderies became factories filled with women inhabiting row after row of sewing racks. These women sewed signatures on cords set into saw cuts on the spine of the textblock. Rather than having to loop the thread around a raised cord, or in a much more intricate pattern around a double raised cord, the thread simply passed behind the now set-in cord. This was much quicker (and, quite arguably, vastly inferior) to the old process. The cords still extended beyond the spine onto the boards. However, by this point in time the importance of lacing the cords to the boards seems to have been forgotten. These books were now casebound, the cover was made independently and attached to the book in a final step rather than built onto the spine/board structure as had been done previously. When case and textblock came together in the final step of the process, the cords were simply splayed out and glued down onto the board between the board and the pastedown. If any, spine reinforcement was minimal (often just a paper lining) and did not extend onto the boards.

    The mass produced nineteenth century book represents a transition between the modern, machine sewn book and the previous tradition of handcrafted books. The cords were now vestigial appendages. The splayed, glued down connection to the boards contributed little to the actual attachment of the boards and a modern system to replace the laced in cords was yet to come. Essentially, the books depended on their endsheets to secure textblock to cover. These books were bound to fail if they received any significant use. When they did fail it most likely occurred at the junction of the pastedown and endleaf. My guess is that these failures led to generations of binders experimenting with various means of reinforcing the hinge between the pastedown and the end leaf. This is still the principal mode of attack for the novice who attempts to replace a detached cover by taping or by some means re-attaching the cover to the endsheets.

    Somewhere in the transition from ancient to modern, binders  seem to have temporarily lost a key concept: the boards were secured by the cords which were secured to the spine of the textblock. The boards (and the consequent cover) were never secured by the endsheets. The role of the endsheet/pastedown assembly was either aesthetic or a means of controlling warpage in paper based boards. Many early wood board bindings simply omitted them. The endsheets played little or no role in securing the cover to the textblock. Attempts to give the endsheets a structural role represent an evolutionary dead end in binding (there are minor exceptions here - the quarter-joint binding being one). The introduction of the cloth super that extended across the spine and onto the boards reasserted the original principal that a book's boards (and therefore its cover) are attached via the spine. Whereas the early book may have been sewn on anything from two to a dozen or more heavy cords that reached across to the boards, the modern super-reinforced binding has literally hundreds of tiny cords (the threads in the super) that secure adhesively to the spine and reach over to secure the boards.

    A brief aside: the emphasis on endsheets as a structural component may have been aided by the concept of thongs and cords as sewing supports rather than as a means of board and cover attachment. Eliminate the boards and the concept of "sewing supports" eludes me as I cannot see exactly what is being supported (that wasn't already supported) or what sewing or leaf attachment problem was solved by the addition of cords. I believe that early codex bindings had neither hard covers nor cords but were sewn with a chain stitch connecting one signature to the next. My unsupported speculation is that the evolution of sewn-on cords and thongs was directly related to attempts to figure out how to get those darn boards (or even soft covers) to stay attached to the textblock. The painstaking manner in which many of these cords or thongs are secured to the boards indicate that their use as a means of board attachment was considered significant. Their effect on the sewing or leaf attachment was incidental to this quest. Furthermore, the addition of raised cords and thongs may have strengthened the textblock but only at the high cost of adding tremendous stress to the sewing when the book was actually opened (see Flexible Strength).

    So, what is the significance of all this? It is that a book's cover or case is secured via the spine and not the endsheets. Books designed and repaired on this principal have a good chance of success. Books built on the premise that endsheets provide a structural role essentially transfer the stress of board attachment from the spine, a very secure base, to whichever page follows the reinforced endsheets, a very weak link. If the endsheets have cloth reinforced hinges and are sewn on as signatures at the beginning and end of the textblock, this may pull the inter-signature sewing into play as a support for the cover. This may not be desirable as it can add stress to the sewing that is better taken on by the super. However, if the book's spine is properly reinforced with a strong super the inter-signature sewing is no longer structurally in play and the reinforcement of the endsheets becomes superficial.

    And as an afterthought, what makes a good super? How well a super does its job depends on its strength and how well it adheres to the spine and the boards. The super should be supple enough when wet with adhesive to conform to minor irregularities in the the spine. It should breathe sufficiently that air can pass through it, preventing air bubbles from forming underneath, and it should not delaminate when subjected to the normal motion of a book's spine. Good adhesion depends on two factors, the adhesibility of the mating materials and the actual surface area available for adhesion. Given any two materials the greater the surface area being glued the more difficult they will be to separate.

    Given these criteria we can evaluate different materials. Paper lacks tear strength, is not breathable, does not conform to irregularities in the spine and delaminates with the normal flexing of the spine. Tyvek has tear strength but shares all of paper's other shortcomings. Gauze or cheesecloth supers come in many different flavors ranging from from very weak to very strong. Their use in glue is frequently compared to steel reinforced concrete. I consider this analogy invalid and misleading as the properties of a flexible bookbinding glue are in no way analogous to concrete nor are the structural expectations of a spine in anyway similar to that of reinforced concrete. Whether weak or strong, all cheesecloth type materials offer less surface area for adhesion than cloths more fully woven. This absence of surface area is most telling not on the spine but where board meets cloth. Cheesecloth supers tend to readily delaminate not from a book's spine but from the hinge and board areas. Some cheesecloth supers also appear to be made of synthetic materials that do not adhere well. A strong, tightly woven (but not so tight it lacks breathability) non-synthetic cloth makes the best super.

    My personal preference is a cotton cloth with a 68 x 68 thread count cloth sold by Library Binding Service ( There are probably similar cloths sold elsewhere - this is simply the company I dealt with. If the spine is going to be reshaped after glue-up, which is the case with fan-glued bindings that will be rounded, the super must be capable of stretching. This requires a cloth specially designed to do so. The super I normally use will split if subjected to rounding. LBS, as well as other bindery supply houses sell stretch lining materials similar to the cotton backlining. However, unless a particular binding requires a stretch cloth there is no reason to use one.

To be continued in Reflections - part 2- Adhesive book structures

Main Menu  


Reflections -part 2