Begin with the object, not the idea. A bound volume of blank sheets. At the front, two facing pages ruled into a grid: twenty letters of the Latin alphabet down one axis, five vowels across the other, one hundred fields in all, with a special dispensation for Q, which in Latin is always followed by u. In each field, the writer enters not words but page numbers. That is the whole apparatus. Everything else — the excerpts, the headings, the reading — happens elsewhere in the book, and is reachable only through this grid.
This is the method John Locke described in a letter to his French correspondent Nicolas Toinard, published in the Bibliothèque universelle et historique in 1686 as Méthode nouvelle de dresser des recueils and issued in English in 1706 as A New Method of Making Common-Place-Books. The 1706 volume, as the Internet Archive’s catalogue record describes it, opens with a double-page table headed “The index,” printed in red and black, before the letter to Toinard begins. The index is not an illustration of the method. It is the method, or at least the part of it that made the rest work.
What follows is an argument about what that grid does to the unit we call an extract. The commonplace book is usually treated as a treasury of quotations, a memory aid, a store of eloquence. Locke’s design points somewhere else: toward retrieval. And that shift matters beyond the seventeenth century, because literary criticism still works with extracts, and still tends to treat them as transparent samples of a whole rather than as objects filed under a scheme.
The received view: commonplacing as supplemental memory
The standard account is not wrong, only incomplete. Commonplacing was a pedagogical practice with deep humanist roots. As Michael Stolberg describes it in Early Science and Medicine, schoolboys following Erasmus and other humanists were instructed to excerpt useful passages from approved authors and arrange them under “topics,” “headings” or “places”; these common places served as models of elegant Latin and as a source of copia, the verbal abundance that enriched speech and prose. The genre’s purpose was rhetorical. You gathered in order to speak and write well.
The memory framing is equally well attested. The Public Domain Review, introducing a digitised copy of the 1706 edition, quotes Jonathan Swift’s 1721 Letter of Advice to a Young Poet: a commonplace book is something “a provident poet cannot subsist without, for this proverbial reason, that great wits have short memories.” Swift’s phrase for the notebook is “supplemental memory.” The book stands in for what the mind loses.
Both framings — copia and memory — treat the extract as a unit of value. The passage is worth keeping because it is eloquent, or because it is memorable, or both. The book is a container for such passages, and its quality is judged by what it holds.
What Locke’s design actually does
Locke’s method reorganises the container around a different question: not what is worth keeping, but how will I find it again. The mechanics are worth stating precisely, because the argument depends on them.
According to the Public Domain Review’s account, the writer begins with blank sheets bound into a volume and a two-page index at the front listing each letter of the Latin alphabet beside each vowel — only the possible combinations, hence just “Q.u.” When a passage is to be entered, the writer first chooses “a proper head.” Suppose the head is Epistola. The writer looks to the index for E and the following vowel, i. If a number appears in the E.i field, it directs the writer to the page reserved for words beginning with E whose first vowel after the initial letter is I; the entry goes there. If the field is empty, a new page is started and its number entered.
Two design choices follow from this, and both are consequential.
First, the heading is chosen at the moment of note-taking, not fixed in advance. Stolberg stresses this: there is no pre-established order, and the writer decides on “some important and essential word to the matter in hand,” preferably in Latin, as each entry is made. The book’s structure is generated by the act of reading, not imposed before it.
Second, the index records page numbers only. It does not list the headings themselves. The grid is a map of addresses, not a table of contents. To use the book you must know the rule — initial letter plus first following vowel — and apply it; the index will not tell you what is in the volume, only where a class of things would be if it were there.
The economy of this is easy to miss. The Public Domain Review notes that the letter-plus-vowel scheme allowed a great range of topics to be located through a two-page index, as opposed to simple alphabetical arrangement, and that in the main notes area it meant less wasted paper than a traditional alphabet-based model in which each letter was preassigned a number of pages — while offering more structure than a purely chronological sequence. Stolberg makes the same point in stronger terms: compared with established approaches, Locke’s method offered maximum flexibility while facilitating later retrieval and minimising waste of space and paper.
What Locke did not invent
Here the argument needs a correction, and Stolberg supplies it. It is tempting to present Locke as the inventor of indexed note-taking. He was not. Stolberg identifies three major approaches to commonplacing already practised by physicians and scholars before Locke: the systematic or textbook approach, the alphabetical approach, and the sequential or index-based approach. Locke experimented with these in his own notebooks — over two dozen of which survive, the large majority devoted almost entirely to natural philosophy and above all medicine — and his published method is a refinement of existing practice, not a creation from nothing.
Stolberg is explicit that Richard Yeo, in the most detailed earlier analysis, wrongly credits Locke with inventing features that were already common by his time. The honest description is narrower and more interesting: Locke formalised a retrieval scheme, published it, and made it portable. The Méthode nouvelle appeared in 1686; apart from a few earlier poems, Stolberg notes, it was the fifty-four-year-old Locke’s first publication. A working method became a printed instrument.
There is a dating wrinkle worth flagging rather than smoothing over. The Public Domain Review describes the method as formalised in a 1685 publication; Stolberg dates the Bibliothèque universelle appearance to 1686, and the Internet Archive record describes the 1706 English volume as a translation of the Méthode nouvelle originally published in the Bibliothèque universelle et historique de l’année 1686. The discrepancy is minor and probably reflects the difference between composition, licensing and printed issue, but it should be stated as the sources give it rather than resolved by assertion.
The extract becomes a unit of retrieval
Now the interpretive claim, which is mine and should be weighed as such.
In the copia model, the extract is a unit of eloquence. Its value is intrinsic to the passage: this sentence is worth having because it is well made or well said. In Locke’s design, the extract is a unit of retrieval. Its value is relational: this passage is worth entering because it can be addressed, and it can be addressed because a heading has been chosen that maps onto a field in a two-page grid. A note that cannot be addressed is, in practical terms, not yet a commonplace entry. It is a note.
This is legible in the design choices without any claim about what individual readers thought or felt. The heading chosen at the moment of entry; the index that records addresses rather than titles; the deliberate tolerance for heterogeneous material on the same page — all of these serve lookup rather than recitation. Stolberg’s account of the method’s fitness makes the point sociologically: it suited physicians and natural philosophers interested in the infinite variety of natural particulars rather than in elegant quotes on a limited set of classical topics. The scheme was built for a reader who expected to return, not to perform.
Robert Darnton’s description, quoted by the Public Domain Review, captures the resulting texture: commonplace users “broke texts into fragments and assembled them into new patterns by transcribing them in different sections of their notebook.” Fragmentation is not a by-product here. It is the operation the index makes possible. You break the text because you intend to reassemble it by address.
Why this matters for literary criticism
Critics quote. We treat the quotation as a sample of the whole — evidence that the text says this, in this way, at this point. The commonplace book suggests a different description of what quotation is: a retrieval act performed under an implicit filing scheme.
Consider a thought experiment, offered as a thought experiment and not as history. Two scholars read the same book and make the same excerpts. One files them under themes; the other files them under dates of composition. Years later, asked what the book is about, the first can retrieve passages by topic and the second by sequence. The difference in what they can find is a property of the filing scheme, not of the reading. Neither scholar read better. They built different indexes.
Locke’s grid makes this visible because it is so explicit. The two-page index is a confession that the book’s usefulness depends on a scheme the reader brings, and that the scheme can be inspected, criticised and revised. Most critical quotation operates with the scheme invisible — naturalised as “the relevant passage” or “a representative example.” The commonplace book tradition is a reminder that relevance is produced by a filing decision, and that filing decisions can be argued about.
This is not a counsel of despair about quotation. It is a counsel of documentation. If the extract is a retrieval act, then the critic owes the reader the scheme: why this passage, filed under this heading, retrieved for this argument. Slow reading, in this sense, is not merely reading slowly. It is reading with the index in view.
The limits of the paper technology
What the sources establish is a design and its stated rationale. What they do not establish is what happened in anyone’s head.
Stolberg discusses the potential epistemic impact of commonplacing on early modern medicine and natural philosophy — its importance for contemporary debates about species and disease entities, and for the emergence of the notion of “facts.” That is a discussion of potential impact, framed as such. It is not a demonstrated causal rule, and it should not be quoted as one. The design of the index tells us what the method was built to do; it does not tell us that users became more empirical, more sceptical, or more systematic in their thinking. The instrument constrains and affords; it does not determine.
Nor does the design tell us about Locke’s own memory or reading experience. The Public Domain Review’s framing — Locke reacting to Pascal’s prodigious recall by attempting to simulate it “not in the mind, but upon the page” — is a useful way into the material, but it is a framing, not a finding. The grid is evidence of a problem recognised and a solution designed. It is not evidence of a psychology.
There is also a gap in the record that any fuller treatment would need to close. Stolberg notes that Locke’s commonplace books have been mined above all for their contents, while the paper technologies he applied have drawn relatively little attention from historians. Item-level study of particular volumes — how a given notebook actually used the grid, where it broke down, what was entered and what was left out — would test the interpretation offered here against the surviving objects. That work is possible; the manuscripts survive in quantity. It is simply not what this essay has done.
What the index teaches
The commonplace book is usually described as a technology of memory. Locke’s contribution is better described as a technology of address. The two-page grid does not help you remember what you have read; it helps you find it again, provided you chose a heading that maps onto a field. That is a narrower achievement than the legend suggests, and a more useful one.
For critics, the lesson is not that we should keep commonplace books, though we might. It is that every extract we quote has an address, whether or not we disclose it. Locke’s index makes the address visible, arguable, and revisable. Criticism that hides its index is not more natural. It is merely less documented.
Questions readers ask
Was Locke the inventor of the indexed commonplace book? No. Stolberg identifies systematic, alphabetical and sequential approaches already in use among physicians and scholars before Locke, and explicitly disputes crediting him with features that were common by his time. Locke’s contribution was a particular refinement — the letter-plus-first-vowel index — and its publication as a portable method.
When was the method first published? The Méthode nouvelle de dresser des recueils appeared in the Bibliothèque universelle et historique in 1686, according to Stolberg; the Public Domain Review gives 1685 for the formalisation. The English translation, A New Method of Making Common-Place-Books, was published in London in 1706. Present the dating as the sources give it.
How does the index work in practice? The writer chooses a heading, takes its initial letter and first following vowel, finds the corresponding field in the two-page grid, and follows the page number recorded there — or starts a new page and records its number. The index contains addresses, not headings.
Does the method tell us anything about Locke’s own reading? Not directly. The design shows a problem recognised and a solution built. Claims about Locke’s memory or mental habits go beyond what the sources establish.
Why should literary critics care about a seventeenth-century filing scheme? Because criticism runs on extracts, and every extract is retrieved under some scheme. Making the scheme visible — as Locke’s grid does — turns an invisible filing decision into an arguable one. That is the beginning of rigour, not the end of it.
