A theory of book publishing: synopsis
Anthony Haynes writes: Over the last few weeks I have published a theory of book publishing in the form of a series of blog posts. The post below collates those previous posts.
When I set up the ‘Towards a theory of publishing‘ page above, I wrote that I was doing so as a substitute for the more extended, rounded, treatment that I could see no immediate prospect of writing. I’m now going to make an attempt at such a treatment, as a result of two catalysts:
- reflecting on Michael Bhaskar’s theory outlined in The Content Machine, reviewed here on 12 November and 19 November last year;
- discovering, when scrolling through my draft posts for this blog, that some years ago, working for a couple of evenings in a cafe in Newcastle upon Tyne, I had in fact made a start on the task.
I had forgotten.
I plan to publish the theory in the form of a sequence of five weekly posts, each published on a Thursday — beginning on 14 January.
Three qualifications before the posts begin:
- the theory should properly be called a ‘pragmatic theory’. There are two reasons for this. First, whereas Bhaskar was concerned to develop a theory that suited all historical situations, I’m really concerned only with the present and foreseeable future;
- I’m more concerned than Bhaskar to derive implications for practice. Though I don’t think a theory must necessarily yield such implications — there’s nothing wrong with pure theory — temperamentally I prefer theories that can in some ways function as tools
- though the theory I’ll be posting does provide a more extended and rounded treatment than the ‘Towards a theory page’, it will remain far from comprehensive.
Overall, a more accurate title for the series of posts would be ‘A sketch for a pragmatic theory of book publishing’. But that sounds too cumbersome to use.
Act I: The basic situation
Commissioning content can happen in at least three ways. (I) An author might have an idea and propose it to a publisher. (II) The publisher might have the idea and propose it to an author. Or (III) the author and publisher might have a creative conversation out of which emerges a proposal that they have co-created.
For the sake of simplicity, the theory outlined on the current series of posts will assume (I).
Let us imagine that there are a number of publishing houses to which the author might send the proposal. A variety of outcomes might ensue, depending on which publisher the author selects. For example:
- Publisher A might reject the proposal.
- Publisher B might feel the proposal isn’t suitable in its current form, but will seek to negotiate some changes in the specification, with a view to then commissioning it.
- Publisher C might accept the proposal and decide to publish it in a certain format — for example, in print only (for the first edition, at least), in hardback.
- Publisher D might accept the proposal but decide to publish it with a different strategy — for example, simultaneously in paperback and as an ebook.
Further permutations are possible. For example, Publisher A might, on a different occasion, accept the proposal. Or some publisher might reject the proposal but, attracted by the author, seek to explore alternative ideas for books. Or the author might decide to self-publish. There are too, many possible publishing strategies.
But, as a way of getting started, I wish to keep things simple. From the point of view of theory, nothing hangs on the specific outcomes identified above. The point that matters is simply a single proposal can have a variety of potential outcomes.
None of the above will be in any way news to people who work in book publishing. But to those who don’t — including many prospective authors — this can indeed be news.
And it can, in addition, be the source of perplexity. In each case, it’s the same proposal: how come the variety of outcomes?
Daily life in publishing doesn’t encourage reflection on this point. There’s too much else to do. Texts need editing, covers need designing, marketing collateral needs producing, orders need fulfilling … and so on. Authors’ perplexity, to the point it is considered at all, may be put down to their naiveté.
The authors, meanwhile, might put the situation down to irrationality on the part of publishers. To an extent, they will be right: irrationality certainly plays a role in publishers’ decisions and the ways in which does so are of theoretic interest. But let us focus here on the opposite case — that is, where differences in outcome are the result of rational behaviour on the part of publishers. That will seem all the more perplexing and is, I suggest, more intriguing.
So, rather than dismiss it, I propose to take the question ‘How come the variety of outcomes?’ as the basic problem for a theory of book publishing. My contention is that formulating an answer will help to explain much else in book publishing — and possibly, by extension, in the creative industries more generally.
Act II: Publishing as a language
In The merchants of culture, John B. Thompson likens the workings of the book publishing industry to a ‘grammar’. Like Thompson, I here adopt a linguistic metaphor. But I think ‘grammar’ is too restrictive. For the purposes of lateral thinking, we need to bring in other aspects of language, such as syntax, lexis, semantics, and pragmatics.
Consider the following sentence (commonly used for typing practice):
(1) A quick brown fox jumps over the lazy dog
And some possible variants:
(2) A quick brown fox jumps athletically over the lazy dog
(3) Fox jumps over dog
(4) A quick brown hat-stand jumps over the lazy translation
(5) Fox a quick brown fox jumps the dog lazy over
(6) A quick the lazy
Though (1) sounds rather odd, it is a perfectly good sentence. It is grammatical (as a subject-verb-object sentence) and it is meaningful. It’s even rather clever – not only because it very efficiently incorporates all the letters of the alphabet, but also because it can be remembered in the form of a mental image.
(2) is also perfectly good. In some ways it is less efficient, but it is also richer: the inclusion of an adverb makes for a more complete treatment.
(3) is good too: it might form a (boring) newspaper headline or telegram. It’s efficient, but impoverished.
(4) makes sense grammatically – the subject-verb-object structure remains. But the change in lexis has rendered the semantics nonsensical.
(5) is a bad sentence — or, rather, no sentence at all. I think of it as an Eric Morecambe sentence: it contains all the right words, but ‘not necessarily in the right order’.
(6) is too truncated to form a sentence.
I suggest that we can draw an analogy with publishing. Each component of the publishing process can be thought of as a component in a sentence (or, indeed, non-sentence). Such components include: commissioning and acquisition; contracting and gaining permissions; development editing; copy-editing; proofreading; indexing; translation; text design and typesetting; artwork and photography; cover design and production; procurement; printing and reprinting; binding; ebook conversion; providing bibliographic data; marketing, publicity and promotion; selling (copies, rights, and permissions); warehousing, distribution, and fulfilment; archiving; paying royalties; remaindering; and managing returns. There are also numerous less industry-specific components — financing, accounting, reporting, recruiting, employing, training, managing, formulating corporate strategy, and so on.
Not all these components need be present. I’ve heard one Managing Director tell me, ‘We don’t have a marketing department’, whilst another told me, ‘We don’t do publicity’. In each case, the missing component functioned as, in effect, the optional adverb (‘athletically’) that differentiated #2 from #1.
But if too many of the components are omitted, we would end up with a malfunctioning or even non-existent sentence (#5 or #6).
It is tempting to seek to equate each component with a particular part of speech, But that is a matter for parlour games rather than for theory. In my experience, each department likes to think of itself as the main verb.
From the point of view of theory, the key points are:
- a minimal set of components is required;
- the components that are present need to be aligned.
Where this is the case, the publishing house equates to #1-3, depending on the comprehensiveness of the service.
Where the components are not (well) aligned, we end up with, at best, #4.
What does alignment mean here? ‘Alignment’ refers to the question of how well the various components of a publishing house fit together. Take the example of textbook publishing in higher education. The text and figures may be developed to suit existing educational courses (taking account such things as educational level, pedagogical practices, and curricular and assessment requirements); the marketing may be targeted at the lecturers who might adopt the book (they will, for example, likely be offered inspection copies), plus the collection librarians in the same institutions; and warehouse stocks will be reviewed and replenished prior to adoption season (typically, at the start of the academic year).
In such a case, the components are well aligned. Here the whole (the success of the operation) will at least equal the sum of the parts. But supposing some component is missing? For example, the marketing department lacks the ability to contact potential adopters (because, for example, they lack the necessary contact lists or do not have the personnel to exploit them) or has not created an effective mechanism for offering and providing inspection copies. In this case, even if the other components are excellent (the development editor, for example, has helped the authorial team to produce a text ideally suited to the market) — and even if the marketing department itself is good at those things it does do (producing pages on the website, for example), there will be underperformance: the whole will be less than the sum of the parts.
All this might sound obvious. In which case, one might think such misalignment is readily avoided. Whilst such misalignment is theoretically possible, does it occur in practice? And, if so, does it ever persist?
The answer to both question is, ‘Yes, frequently!’ In my career I have seen, and continue to see, misalignment on a regular basis. To take, by way of illustration, one example, I helped an author obtain a contract for a scholarly monograph. The book was commissioned, contracted, and written on that basis. The sales forecast was for a few hundred copies in hardback, at a price aimed at institutional libraries. The design department then produced a beautiful four-colour dust jacket, complete with a photograph for which they bought copyright permission. I dare say that the designer was pleased with her/his work: everyone agreed it looked attractive and, I’m pleased to say, the author was especially chuffed. The book would, no doubt, look great in a bookshop — would do so, if only bookshops stocked expensive hardbacks. Doubtless, too, it would look attractive to individual readers — except they don’t buy expensive hardbacks. Libraries do buy such things and did in fact buy this book in pleasing numbers. They buy such books through sales plans provided through library suppliers, based on bibliographic data. If they receive a copy with its dust jacket intact, they will remove the jacket before shelving the book.
Many scenarios produce misalignment. For example:
a) a company might suffer from silo-isation, where each department does its own thing and no department regards inter-departmental co-ordination as its own responsibility. (In my experience – though how typical this is, I’m not sure – when it comes to destroying value through silo-isation-based misalignment, university presses are unsurpassed.)
b) a company might have insufficiently robust ways of coding projects, What begins life, at commissioning stage, as a textbook might not be signalled to the marketing and sales department as such a thing. It may then receive a marketing and sales treatment more appropriate to some other genre.
c) a company might acquire lists from another company. Some of the titles acquired may not align with the acquiring company’s processes.
d) an organisation might publish content, whilst lacking specialist knowledge of publishing. Such organisations can include charities, think tanks, and public sector bodies.
From this conception of the publishing process as akin to sentence formation, several practical corollaries follow. In particular:
- The process may be improved by improving a specific component. To continue the analogy: just as we might make the verb in #1 more concise by replacing the two words ‘jumps over’ with ‘hurdled’, so we might improve the marketing department by upgrading their software.
- The process might be improved by adding components. For example, a small company might lack the means to sell translation rights. As it grows, it might appoint a rights manager, establish a rights database, and perhaps visit trade fairs. In the linguistic analogy, this equates to moving from #3 to #1 (or #1 to #2).
- But the process might also be improved by improving, not the components themselves, but the alignment between them (moving, as it were from #4 to #1). To give two positive examples here: the MD who told me that his company didn’t have a publicity department in fact ran a well aligned company (it published textbooks, which benefit little from publicity); likewise with the MD whose company had no marketing department — it published consumer books that found readers through such methods as media coverage.
- Given that alignment is far from automatic, a company can benefit from, as it were, knowing what kind of sentence it is. Once this kind of self-knowledge is achieved, the various components can work in collaboration. If, for example, the commissioning editor knows what kinds of books the marketing and sales department can succeed with, the editor can commission accordingly — rather than commission other types of books and then blame colleagues when they don’t succeed. We can say that accurate prediction and efficient allocation of resources depends on understanding the sentence that the company constitutes.
- Wherever a company publishes more than one sort of book (many companies, after all, consist each of a set of parallel sentences), an effective means is required for coding each book. Failure to code books accurately (for example, through the way that records are entered on a database) results in, in effect, trying to get hat-stands to jump over dogs (or, worse, over translations).
Act III: Accounting for diversity
This series of posts is focused on diversity between publishing houses. Why is it that the same book proposal will achieve different outcomes, depending on which houses it is sent to? In the previous post I argued that each publishing house could be thought of as constituting a distinct sentence (or set of sentences). But this raises the question, what accounts for the differences between those sentences — why don’t publishing houses constitute identical sentences?
Empirically, there are many reasons. The staple of literature about publishing is the company history and each history accounts for the distinctiveness of its subject. But here I’m interested in the typical reasons for diversity — what in general accounts for the diversity of outcomes that would follow from the same book proposal?
I suggest that there are two factors. First, each house constitutes a bundle of agreements with stakeholders. The most long-term and stable of these are formal contracts. Alongside these are shorter term or less formal agreements, for example in the form of purchase orders and letters of agreement. And then there are still less formal, though sometimes very long-term, relationships in the form of networks. All three kinds are relevant, though my focus here is on the least flexible of these, namely contracts.
These agreements define a publishing house’s relationships with its customers, suppliers, employees, and associates. On these agreements will depend what ensues from a specific publishing process.
For example, one British publishing house might have little capacity to publish directly in North America, where they prefer to co-publish. Perhaps too they sell translation rights to other export markets via a third-part agent. The house might have an effective, though indirect, relationship with the domestic trade through their distributors, who supply direct to shops and also to wholesalers. Perhaps they place most of their printing with one printer, but use a different firm for short-run and on-demand printing. And they tend to recruit staff through a specialist agency.
Another house might sell copies direct in North America. And they might, say, sell translation rights direct, though the effectiveness of this arrangement might be limited by the number of book fairs they attend — perhaps they go only to London and Frankfurt. Perhaps they have strong relationships with international library suppliers, including suppliers of ebooks. They have little relationship with national mass media and rarely get their books reviewed, other than in specialist publications. Perhaps they shop around for printers on an ad hoc relationship. They often promote from in and, for recruitment, rely on networking supplementing by advertising in the trade press. Some of their employees started as interns.
Let us suppose that an author has a book proposal that both houses would, potentially, be interested in. How the book fares will vary as a result of the differences in the types of arrangements outlined above. Perhaps the former house sells a higher number of copies in America, but because of co-publishing discounts achieves no greater royalty for the author. Perhaps that house does well with co-publishing agreements in the Commonwealth too and sells the Chinese language rights. Print copies look identical and bear a close relationship to the house’s other publications. The author is chuffed to see reviews in the mainstream press.
Alternatively, perhaps the latter house picks up a few deals for Spanish language rights and assorted eastern European territories. Through suppliers, it does some tidy business with libraries in various territories. Because of this market, it’s able to maintain a higher cover price and also to sell an ebook edition, perhaps over a period of years, at prices well above the Kindle-style bargain basement price points. The author receives a few enquiries about her/his consultancy services as a result of trade press coverage. The person who ends up marketing the book was, at contract stage, the editorial assistant.
And so on. The outcomes — for the publishers, the author, for consumers, and for the book itself — will vary according to the agreements in place. Which house is preferable will depend on the context.
A second factor is corporate memory. Let us consider again the former publishing house above. Perhaps there are certain kinds of books that that house ‘gets’: that is, it understands them and knows how to succeed with them. It is difficult to say where exactly this kind of corporate history resides. In fact, it will be dispersed amongst a number of sites — for example, the editor’s head, the sales data, the lists of reviews copies, and the design production department’s archive. If another book of the kind that it ‘gets’ — the latest in a series on military history, say — it’s likely to do an efficient job. The amount of thinking and the number of decisions to be made will be minimised, as will (as a result) the likelihood of misunderstandings or errors of judgement.
But what if a new kind of book — new to that house, at least — comes along? It may do fine: perhaps the novelty energises the staff. But the risk of misjudgements is certainly higher. So too is the likelihood that the house will somewhat reduce the degree of novelty by treating the new book as much like the usual kind as possible.
For example, an academic publisher that attempts to publish occasional trade (i.e., consumer) books may use a largish page format, not because that is optimal for the book in question, but because – well, that’s just what they do.
The combined effects of these two factors — bundles of agreements, plus corporate memory — can be strong in terms of maintaining differences between houses. Indeed, the factors can work in tandem, each constraining the other.
One might argue that movement in staff between houses would serve to reduce these differences. Publishing, like many creative industries, exhibits some degree of geographical concentration. In the UK, the main centres are London, the area around the M25 (the outer ring road for London), and Oxford. The concentration facilitates staff movement between companies. So let us suppose that the former house above recruits a member of staff from the latter. This may indeed result in a change in the new employer’s practices. Perhaps, for example, they experiment with a different printer: the recruit has used them at her/his previous company and reports that they did a good job.
This effect is not, however, as strong as one might expect. After a while, usually quite a short while, there is a tendency for recruits to go native — that is, to adopt the mindset of the employer they have moved to. And so the effects, of agreements and of corporate history, can persist over time.
ACT IV: The question of self-publishing
The advent of (a) digital printing, enabling short-run printing and even print on demand, and (b) ebooks has sharpened the question of what it is that publishers do and how they might add value. Disintermediation (here defined as the elimination of the publisher as an intermediary between author and public) has become more attractive than it was.
The accounting of the benefits and costs of self-publishing relative to the classical model of publishing are more complex than is commonly recognised. Typically over-looked factors are the opportunity costs of self-publishing — especially regarding the value of the author’s time.
I have outlined these considerations in a previous post — ‘Self-publishing: why (or why not) and how)‘, 10 November 2012 — and so won’t repeat them here. Suffice it to say that we can summarise the sources of value added by the publisher under two headings: (a) the co-ordination of services and (b) branding.
Co-ordination of services
Many services are outsourced by publishers. For example, publishers might use freelancers for such services as copy-editing, proofreading, and indexing. In fact, in principle virtually everything that a publisher does can be outsourced: it is difficult to identify an essence that must be kept in-house. Hence the attraction of self-publishing: if a service can be outsourced by a publisher, then in principle it may be outsourced by a self-publisher.
But publishing is more than the summation of such services. Publishing does not equal editing + typesetting + printing, etc. This is because, in addition to such services, those services require co-ordination. It is in this co-ordination, the demands of which are sometimes overlooked by prospective self-publishers, that much of whatever value-added there may be resides.
‘Co-ordination’ here means, in part, just management – for example, placing an order with a printer. But it also means something like ‘bringing into alignment’. For example, the publisher might have a vision for a book and then select suppliers (designer, typesetter, indexer, etc.) well suited to the type of work in hand (rather than simply the cheapest) and will provide them with briefs designed to communicate the vision. In this way we may say that there is a holistic element to co-ordination – a holistic element that requires publishing intelligence, regardless of who is doing the publishing. Thus co-ordination — ‘holistic co-ordination’, if you like — should have a central place in any theory of book publishing.
Even co-ordination can, at least to some extent, be outsourced. For example, we have on occasion used the excellent Out of House Publishing to manage projects for us: they have sourced the editing, proofreading, typesetting, printing and binding for us and managed the authors during the process. Even there, however, we retained a share in the holistic element of co-ordination: we formulated and communicated the brief, which they then implemented.
Publishers also, except in cases of contract publishing, provide a brand. Some of these brands are recognised by consumers. For example, Mills & Boon, Haynes car manuals, and Penguin. True, most publishing brands aren’t well recognised amongst the public. (For example, how many people reading the Booker Prize winner can tell you who published it?) Brands do however mean something within the supply chain. Brands such as, say, Atlantic or Sage do mean something to businesses such as retailers, wholesalers, and library suppliers.
In general, then, there are two distinctive publishing elements: (holistic) co-ordination and branding. I’m tempted to invent a term — say, ‘publish-icity’ — to encapsulate them.
ACT V: Book publishing as a capitalist pursuit
The development of the printing press was a thoroughly capitalist affair. The plant required was expensive: financing it required a large lump sum. Because, in the early days, publishing and printing tended to be combined functions, it follows that publishing too tended to be a capitalist affair.
Over time, the two functions became unbundled: printers printed and publishers outsourced their printing requirements to them. This reduced publishers’ capital requirement.
But traditional printing methods entail set-up costs: to obtain economical unit costs, publishers typically had to order large print runs. To do so, capital was required. In modern times, the sums required have been, in the case of most projects, quite modest. For example, the total cost (defined as payment to printers and other suppliers, such as typesetters, and so excluding staff and office costs) of a typical monograph has always been below £5,000.
Still, that’s not nothing — you can buy a second-hand car or two or three good holidays for that. And, in traditional publishing, there are economies of scale, so publishing houses have typically needed programmes of a certain scale. Publishing a few dozen monographs per year takes the requirement for into six figures.
Digital printing, which enables cost-effective short runs, changes the equation. The threshold capital requirement comes crashing down. Micro-publishing and even self-publishing can be made profitable. Individuals with no more than modest means can publish commercially.
Does this mean that publishing is no longer essentially a capitalist affair?
The answer is no.
Types of capital
The reason for this answer is that, as economic theory has increasingly recognised, capital is not confined to the two types referred to above, namely financial capital and technological capital. There is no consensus on how many types of capital there are — how you categorise capital depends on your purpose — but from the point of view of book publishing theory, I propose the following taxonomy:
- financial capital
- technological capital
- human capital
- social capital
- cultural capital
- natural capital.
I’ve covered (1) and (2) above. What about (3) – (6)?
The theory sketched in the preceding posts has drawn attention to human capital in two respects. First, limitations in human capital impose limitations on publishing: I have argued that the range of publications that each house can manage effectively is limited by the extent of its corporate memory.
Second, good publishing requires ‘publish-icity’ — my ungainly neologism for the ability to ensure that all aspects of the publishing process work in harmony. It is, of course, entirely possible that a self-publisher will possess this quality, but it is by no means necessarily the case: my argument is that this factor has typically gone unrecognised when classical publishing has been weighed against self-publishing.
Obviously, many other forms of human capital are involved — the skills involved in editing and design, for example. But such skills are probably more widespread (though the industry seems short of digital and data analysis skills) or, crucially, easier to procure through outsourcing, than is publish-icity.
My theory involves social capital in one place in particular. I’ve argued that a company’s effectiveness is in part a function of its bundle of agreements with other parties. Some of these agreements do not require social capital. Pretty well any publishing company could, for example, open an account with a print-on-demand service such as Lightning Source. Still, partners willingness to trade — and one’s knowledge of potential partners in the first place — is a function of social capital.
For example, our company has a number of long-standing relationships with suppliers. They include The Running Head (text design and typesetting), Benn Linfield (typesetting and cover design), and Carr Design Studios (cover design). The quality of our publishing depends hugely on the quality of their services — and our enjoyment of publishing lies in part on the professionalism of their dealings. In turn, we’ve always been happy to pass them our business.
In most cases, our suppliers consist of people or companies that I had worked with in corporate publishing. In the case of The Running Head, that is not the case — I came across them through an internet search. Even there, though, our decision to approach and commission them was in part because I knew, or at least knew of, some of their trading partners.
The main way in which my theory has involved cultural capital is through the role assigned to branding. Branding can help to differentiate classical publishing and self-publishing. In some cases, the self-publisher might be a bigger consumer brand than the publisher — for example, where the self-publisher is already recognised as a guru in a particular field. On the other hand, when it comes to B2B branding (the kind of branding most relevant in the supply chain), the classical publisher is likely to have the advantage.
Though discussion of cultural capital in relation to book publishing is quite common, I suspect that most publishers’ cultural capital, beyond the B2B aspect of branding, is in fact rather weak. The cultural capital required is provided more by their content and their authors than by the publishers themselves.
Natural capital refers to the stock of assets (broadly defined) provided by the environment and its ecology. I have not referred to natural capital in this series of posts, though I have elsewhere. In this theory, I will leave the subject of natural capital to an epilogue.
Book publishers’ need to access technological capital has for a very long time been met primarily through outsourcing to printing companies. The requirement for financial capital has been lowered by the advent of short-run digital printing. Cultural capital does not seem to be a bulwark of the industry. But human and social capital remain important sources of competitive advantage.
Book publishing remains a strongly capitalist endeavour — just not in the way that most people think.
In my previous post (‘A theory of book publishing, Act V: publishing as a capitalist pursuit‘) I analysed book publishing in relation to five types of capital requirement:
- financial capital
- technological capital
- human capital
- social capital
- cultural capital
Here I consider a sixth type of capital, namely natural capital.
There’s been much research on the impact of book publishing on the environment. Much of this has centred on the impact of (a) the paper industry and, to a lesser extent, (b) printing.
A particular focus has been the comparative impacts of print and ebooks.
Research to date has undoubtedly proved informative. It has, for example, helped to draw attention to the fact that the environmental impact of the paper industry concerns not only the quantity of forests but also their ecological diversity: extensive monoculture can destroy ecosystems. It has emphasised how environmental impact cannot be reduced to a simple matter of carbon emissions (a reductivism that equates to a monoculture of the mind) — there are such matters as biodiversity and conservation to consider.
Yet the research, and the impact that it has had on debate, has limitations. In particular, there has been some tendency to:
- treat certification of paper (for example, by the Forest Stewardship Council) rather uncritically. Though such certification is helpful, some critics have argued that certification is only as rigorous as the procedures that, in reality, govern it. Some heavily forested parts of the world are not widely regarded as free from crime or corruption;
- present ebooks over-positively: though they may avoid some of the negative impacts of paper and printing, they are not without environmental costs themselves. Think power, servers, and lithium batteries. In addition, the fast rate of technical innovation has led to a high turnover of devices.
- look for technical absolutes (are ebooks better than print books and, if so, by how much?) rather than present environmental impact as relative to producers’ and users’ behaviour.
For example, the environmental impact of print books depends on where they are printed relative to the customer (if there is demand in Australia for a title published in the UK, does the publisher ship copies from the UK or print locally?).
Similarly, we need to examine the extent to which readers recycle their copies, for example through lending or trading through the secondhand market. And by how many people are library copies read?
There’s also been a tendency to ignore certain types of environmental cost. For example, there is the impact of commuting (of publishing stuff, book-buyers, and library users), business travel (transatlantic flights of editors to conferences, for example), and impacts embodied in capital goods such as printing presses.
The ideal would be the provision of a tool to enable each publishing house has to assess its environmental impact. There’s still a way to go.