Things that seem true, written out at length.

On the brief

Note no. 01

A written statement of intent — three hundred words, no more — is worth a hundred moodboards. It forces the argument into the open. It's harder to paper over a disagreement in prose than in pictures. Two studios can look at the same Pinterest board and read opposite things into it; they cannot read opposite things into a sentence that begins "this book must feel as though it has always existed in the reader's hands."

A brief also dates better than a moodboard. The moodboard will, in two years, look like a period. The sentence will still read as a sentence. This matters because the books that make it through to print are never the ones you draw first. The brief is the only continuous document across drafts; it is where the argument lives.

This is not a claim about what any particular studio does. It is a claim about where projects actually go wrong. They go wrong in the language before they go wrong in the drawing. If a brief can't say what the book is for, no amount of drawing will rescue it — the drawings will be unusually lovely and unusually pointless.

In defence of Helvetica

Note no. 02

Helvetica is fine. Most arguments against it are arguments against the way it has been used, which is a different problem. Set at the right size, with the right leading, on the right stock, with something sharper to lean against, Helvetica does its job. The job is to get out of the way. It does that more reliably than most of the typefaces commissioned to replace it.

The real complaint about Helvetica is not about the letters but about the kerning pairs that ship with the system version — a different family, strictly, from any of the Linotype, Neue Haas, or revived digital cuts. Designers who have only ever set system Helvetica are arguing against a ghost. Set the real Helvetica, on paper, and a lot of the objections evaporate.

What remains is a fair argument: Helvetica's neutrality is a particular posture, and the posture has a politics. To choose it is to decline to take a position on the page. Most of the time that's the right call — the book isn't about the type. But when the book is about the type, Helvetica does nothing for it. That is a fault of deployment, not of design.

On margins

Note no. 03

The margin is not empty. It is doing load-bearing work. It holds the page open; it gives the type somewhere to end. A book set too close to its edges reads as anxious. A book with generous outer margins reads as confident — even when the prose is not. This is the cheapest trick in book design and the one most often forgotten.

The golden-ratio rule of thumb (1 : 1.5 : 2 : 3 for inner, top, outer, bottom) is a fine starting place and a poor finishing one. It assumes the book will be held; hardbacks with stiff boards swallow the inner margin into the gutter and need more. Paperbacks that crack at the spine need less on the inside and more everywhere else so the page can breathe without the book being broken.

The outer margin is, in a novel, a thumb rest. Leave room for a thumb. The top margin is where the eye enters — make it the most generous one. The bottom margin catches the weight of the block. If you set all four equal, you will have made a page that looks, at a glance, like it's falling over.

What printers know

Note no. 04

A good printer will tell you, politely, that your file is wrong. A great printer will tell you, politely, that your design is wrong. The distinction is worth paying for. Most of what's beautiful in print history was saved, quietly, at the press — not at the drawing board.

The printer knows what the ink does when it hits the paper and the designer usually does not. Solids spread; fine rules disappear; knock-out type at 6pt fills in. A designer who has set foot in a press room twice will make a hundred fewer mistakes than one who has never gone. It is the single most useful thing a book designer can do, and the one most often substituted by a proofing workflow on a screen that lies in seven colours.

If you can get a printer on the phone before you finalise a page, do. Tell them the paper, the press, the run-length, and what you are trying to achieve. They will tell you which of your choices are going to embarrass you in three weeks. They will do it for free. They would rather have the argument now.

On finishing

Note no. 05

The last ten per cent of a book is where it becomes a book. The binding, the endpapers, the spine, the thing it feels like in the hand at three in the morning. Studios that skimp here are solving a different problem than the one the reader has.

A reader buys a book and then they own it for twenty years. The cover comes off the shelf a thousand times. The spine cracks where the readers like it cracked. The endpapers show scuffs where a careful thumb touched them every night for a winter. If any of these elements is plastic-feeling, machine-numb, or the wrong colour for the book it contains, all the care spent on the text blocks will, over that twenty years, be cancelled out by the irritation of the hand.

The book is a thing-that-is-held before it is a thing-that-is-read. That is the order of things. Finishing is where that order is honoured or ignored.

On the silence between letters

Note no. 06

Kerning is not a fix. It is the thing itself. The job of the designer is to make the spaces between letters read as one continuous silence — not six silences, not a silence with a stumble halfway through. When you cannot see the kerning, the kerning is right.

The temptation is to kern by rule — tighten all T's into all A's, loosen all lowercase k's from their neighbours. Rules get you to 80 per cent. The last 20 is by eye, by hand, by turning the page upside down and looking at it as shape rather than language. A single wrong pair on a title page is visible across a room even though you cannot say what is wrong.

Most retail fonts ship with kerning that is good enough for running text and worse than nothing for display sizes. Above 36pt, assume you will re-kern every pair by hand. Below 14pt, trust the font and move on. This is one of the few reliable rules in the discipline.

Paper is an argument

Note no. 07

The choice of stock is the first thing the reader receives and the last thing most studios decide. This is the wrong way round. Every other decision — type size, leading, margin, finish — is downstream of the paper. Soft uncoated stocks drink ink and make serifs heavier; hard coated stocks repel ink and make hairlines invisible. You cannot set the same typeface at the same size on both papers and get comparable books.

If the paper is an argument — and it always is — a cream, bulky, uncoated stock argues that the book is to be lived with. A bright white, smooth, coated stock argues that the book is to be referenced. A tissuey bible paper argues that the book is to be carried everywhere, forever. Choose which argument you are making. Then design downwards from it.

The grid is a promise

Note no. 08

A grid is a promise made to the reader: things-of-the-same-kind will land in the same place. If it is kept, the reader stops noticing the grid and begins to notice the content. If it is broken, even once, by laziness rather than intent, the reader stops trusting the page.

Breaking the grid intentionally is a separate matter and a legitimate one — it is how you mark emphasis, surprise, the point at which the book steps out from behind its own premise. But break it fifteen times and each break will mean less than the one before, until the grid is not a grid but a suggestion, and the reader has long since stopped paying attention to what you were trying to say.

This is why the grid is conservative in the true sense: conserving. It banks the reader's trust, and each accurate placement is a small deposit. Spend that balance deliberately.

Against the final version

Note no. 09

There is no final version of a book; there is only the version you could no longer keep paying for. Admit this up front and the revisions stop being crises and start being choices. The book is never finished; at some point, it is simply bound.

This does not mean the ending is arbitrary. It means the ending is a decision about where the marginal improvement from another week of work is finally less than the marginal cost of the reader not yet having the book. That is an economic argument, not an aesthetic one. Design decisions that pretend otherwise end up in the bin anyway.

Small studios, large rooms

Note no. 10

The best book-design studios fit in one room. Not because small is beautiful in the Schumacher sense, but because the argument about the book must happen between people who can hear each other without shouting. A studio large enough that it needs a Slack channel has, in some measurable sense, lost the thread.

67b is an imagined studio of one, which is either the natural form of the craft or a dangerous tendency, depending on which Monday you ask. The argument on this page is that small rooms are a condition for books that feel as though they have been made by a hand rather than by a process. The counter-argument is that a hand can be wrong in ways a process cannot be, and that is also true, and that is why the best small studios keep honest friends nearby.