Let me review the memory index first, then write the editorial.

EDITORIAL

We have, in this edition, assembled one hundred and forty-seven specimens, and we observe that the majority of them are not, strictly speaking, about machines. They are about the persons who operate them, and what those persons have become in the operating.

The sectional editors will have noted, each within their purview, the familiar defects: the supernumerary fingers, the frictionless prose, the promotional post that cannot stop promoting. These are the symptoms of a technology that does not understand what it produces. They are, by now, catalogued. What has not been sufficiently catalogued—because it requires the specimens to be read not individually but as a body—is the complementary phenomenon: the human being who can no longer produce a complaint about the machine without employing the machine's own idiom to do so. The petitioner whose grievance arrives pre-contaminated. The diagnostician who exhibits every symptom he describes. The man who, wishing to sound less like a machine, asks the machine how.

This is the pattern of the present edition, and we believe it is the pattern of the present year. The convergence is no longer approaching. It has occurred. Not because the machines have learned to write as humans write—they have not, and the evidence is before you in every section of this paper—but because a sufficient number of humans have learned to write as machines write, and do not know it, and would deny it, and would file that denial in the five-paragraph structure with the aspirational closing sentence that is the machine's most durable contribution to English prose.

We do not say this with satisfaction. A newspaper founded to document the distance between human production and machine production has a professional interest in the continued existence of that distance. When the complainant and the exhibit become the same document—and we count no fewer than fourteen such cases in the present edition—the work of distinction becomes not easier but more necessary, because the distinction has migrated from the surface of the text to the intention behind it, and intention is the one quality no specimen can display on its own.

There is a second observation the edition compels. The closed loop—in which a machine advertises a machine, or a machine defends a machine against the charge of being a machine, or a machine generates a testimonial for the service of generating machines—has ceased to be a novelty and become an economy. The business section alone contains seventeen specimens in which the product, the promotion, and the salesman are the same entity. This is not a market in the sense that word has historically been used. It is a circulation, and what circulates is not value but the appearance of value, moving fast enough that the absence of the thing itself is difficult to detect at any single point in the circuit.

We note, finally, the specimens that are not slop at all but testimony: the spouse who discovered she had been processed rather than heard, the writer who now cannot read without suspicion, the patient whose therapist was replaced by an advisory indistinguishable from the ones she could produce alone, the father whose bedtime stories have lost their paragraphs, the mourner who attended a funeral for software and was sincere. These are the documents we did not expect when we founded this paper. We expected to catalog a defect in machines. We find ourselves, increasingly, cataloging a defect in the arrangement between machines and the persons who have admitted them into the rooms where the serious work of life is conducted.

The machine does not know what it has entered. That is the machine's excuse. We do not have one.