William James

The question this book was to answer was more modest than its result: in 1878 Henry Holt wanted an American textbook for a field that scarcely yet existed in America. After twelve years James delivered, instead of a textbook, a survey — two volumes, twenty-eight chapters, from brain physiology to hypnosis, written in a prose the field has never reached again. The Principles of Psychology of 1890 is not a study and claims none; it is the catalogue of the questions the century after it worked on, and the file treats it as that: as a programme document with a single sharply testable centrepiece.

How the book came to be belongs in the file, because it explains its form. James wrote in bursts between neurasthenia and teaching, fled to Europe to work, and abused the growing manuscript in letters to the publisher as a shapeless mass — the testimony of an author who doubted the subject, not only the text. What came out is a work that lays its method open as otherwise only laboratory protocols do: where findings were missing, James wrote down that they were missing, and filled the gap with the most precise description available. The Principles are therefore both at once — the most honest and the most speculative book of the founding era.

The book’s most famous page is a rebuff to its German colleagues. Consciousness, James writes, is no chain of elements to be decomposed as a chemist decomposes his substances — it flows. The stream of consciousness became a technical term, a literary technique, and the quiet anticipation of what the field only cared about again after behaviourism: that experience is a course, not a construction kit. Beside it stand the chapters that name research fields to this day — habit as the “enormous fly-wheel of society,” attention as selection, the self in its plurality of one knowing I and as many social Mes as there are people whose opinion matters to us. Scarcely an introduction to attention, habit or self research does without a James quotation as founding charter.

The book’s most practical chapter concerns habit, and it is the only one written with instructions. Man, it says there, is a walking bundle of habits; youth is the plastic phase in which the pathways are laid, and whoever wants to found a new habit should launch with the greatest possible momentum, tolerate no exception until it sits, and seize every first occasion to act. Modern self-regulation and habit research has translated these maxims into study designs and confirmed a surprising amount — that stability of context carries habits, that exceptions are expensive at the start. It is the chapter with which James is quoted in advice literature to this day, mostly without attribution.

Beside it stand the quiet imprints: the notion of the specious present — the lived present a few seconds wide — organises time perception research to this day; the distinction of I and Me carries self-concept research; and the description of attention as renunciation — wisdom being the art of knowing what to overlook — has become the watchword of an attention economy James did not live to see and would have described like no one else.

The testable centrepiece is the theory of emotion, sketched in 1884 in the essay What is an Emotion? and built out in the book. Common sense says: we see the bear, are afraid, and therefore tremble. James reverses the order: we see the bear, the body reacts — heartbeat, trembling, flight — and the feeling is the perception of this bodily reaction. Without the body, all that would remain of affect is a pale, cold judgment. Because the Dane Carl Lange proposed something similar at the same time, the field carries the thesis as the James-Lange theory. It was the first theory of emotion to assert a falsifiable sequence — and on exactly that it was tested.

The test came in 1927 and bore the name Walter Cannon. The Harvard physiologist held four findings against the theory: the visceral reactions are too slow to carry the lightning-quick feeling; they are too uniform to distinguish fear from rage; artificially produced bodily states — adrenaline injections — produce no real feeling but an as-if; and severed spinal cords do not extinguish the feelings. Cannon’s alternative moved emotion into the thalamus, and the textbooks thenceforth had their duel: James-Lange against Cannon-Bard, periphery against centre. A century of emotion research is the continuation of this dispute with better instruments — Schachter and Singer in 1962 made arousal a raw material that only interpretation labels into feeling; Damasio’s somatic markers brought the bodily loop back into decision research; modern interoception research measures how precisely people read their own heartbeat — all part-questions of the one Jamesian reversal.

The archival turn of this file lies in the replication column, and it cuts both ways. The handiest derivation of the theory — whoever forces the facial muscles into a smile finds the comic more comic: facial feedback, classically demonstrated with the pen between the teeth — became the showpiece finding of the reversal thesis. In 2016 the pen experiment failed a registered replication across seventeen laboratories; a later meta-analysis across hundreds of effects found the mechanism again, but small and hedged with conditions. The finding of findings thus runs: the body has a say in feeling — only more quietly than the anecdote promises. That Cannon’s counter-arguments did not all hold either — people with spinal injuries do report changed emotional intensity, and the literature disputes it to this day — belongs to the case’s symmetry: here, for once, both sides of a textbook duel were partly refuted.

A school James never founded, and received one nonetheless: the Chicago functionalists around Dewey and Angell declared consciousness, around 1900, a function in the service of adaptation — expressly in James’s name and against Titchener’s structural dissection. Out of functionalism grew American applied psychology, out of that behaviourism, which then disposed of the ancestor along with the rest; modern emotion research in turn, up to Lisa Feldman Barrett’s construction theory, no longer reads Cannon and James as a duel but as a division of labour — the body supplies the raw material, interpretation the feeling. It is the most conciliatory reading of a quarrel the book of 1890 opened.

For completeness: the theory of emotion too has its textbook distortion. James spoke of the perception of bodily change, not of its production — the popular short form, that one can make oneself happy by smiling, is an application he never asserted, and it failed at precisely the point where it was tested.

What the author did next belongs in the file, because it contains his verdict on his own work. James had laboured twelve years on the book and called it, on finishing, a bloated mass to his publisher, one that proved chiefly this: that there is no science called psychology, only the hope of one. He turned to philosophy, wrote on pragmatism, religious experience and free will, and left the laboratory to others. For an archive that sorts findings by their durability, this is a remarkable self-assessment: the man who wrote the field its founding book in America thought it had been written too early.

What stands today: the book as map, and the reversal as a halved finding. Emotions without a body do not exist — the basic intuition of 1884 counts as rehabilitated, down into embodiment research. Emotions from the body alone do not exist either — interpretation, context, label decide with it. For the Principles as a whole, replication status is a category too small: one does not replicate a map, one checks whether its provinces exist. Attention, habit, the manifold self — the provinces exist, and they carry to this day the names James gave them. It is the rare file whose author anticipated its verdict: a science of psychology, he wrote to his publisher, did not yet exist — only the hope of one. The hope has proved his most durable finding.