On March 8, the New York Times admitted to having marginalized women and other groups as it announced the new “Overlooked” category to its obituaries section. Four days later, Susan Goldberg, editor in chief at National Geographic, acknowledged the need to make amends for the magazine’s racist legacy. These important steps, which draw on crucial work of scholars, journalists and editors, are part of an overdue reckoning within print media. But they are also only the tip of the iceberg in mainstream American publishing, especially its more lucrative trade branch.

For decades, the book industry perpetuated similar biases. Its representatives determined the range of topics about which the public could learn and molded the very institution of authorship by excluding ethnic and racial minorities and immigrants.

In recent years, movements like #WeNeedDiverseBooks have begun to redress the damaging consequences of such biased curation. The underrepresented writers. The readers whose realities are absent from book pages. The lack of distribution platforms. Thanks to their efforts, the present is changing rapidly. The past of the gatekeeper-saddled book business, however, is as opaque as ever. We have yet to hear the publishers’ mea culpas or recover the voices lost to prejudice, leaving huge swaths of American writing — and pieces of our past — unread.

The costs of this exclusion are starkly visible in one of the most pivotal events in the 20th century: World War II. The 1940s may have been an “intensely literary decade,” but paper usage withered to 75 percent of prewar levels by 1944. Publishing staff was called up for service. Increasingly with time, books were weaponized — turned into anti-Nazi propaganda. Wartime meant a shrunken market, but one in which ethics would have been all the more essential.

Despite this expectation, ethics were hardly a priority even for the most progressive big publishing houses. One of them was Houghton Mifflin (now Houghton Mifflin Harcourt), based in Boston since 1832, had earned a reputation as a WASPy purveyor of Anglophone crème de la crème: Emerson, Hawthorne and Browning. In contrast to its worldlier competitors, especially Alfred A. Knopf or the just-founded Pantheon Books, Houghton Mifflin had long honed its all-American brand. In wartime, it aspired to be “completely root-conscious,” as an editor put it in 1943.

But whose roots were at stake?

By most standards, Adolf Hitler’s “Mein Kampf” would not have fit the “all-American” bill. And yet, as historians James and Patience Barnes write, the press shot to prominence after printing the book’s first abridged American translation in 1933, despite several Jewish groups’ complaints.

As though to prove the worriers wrong, the book sold poorly, though only until Austria’s annexation in March 1938. Suddenly aware of its market significance, Houghton Mifflin fought competitors Reynal & Hitchcock and Stackpole Sons Inc. for the rights to the unabridged version. With support from Hitler’s German publisher, Eher Verlag, it won, establishing itself as the exclusive copyright holder in the United States.

In contrast to Reynal and Stackpole, which had used the profits it earned before Houghton Mifflin won exclusive rights to assist refugee organizations, Houghton Mifflin shared the royalties with Eher Verlag until America’s entry into World War II in December 1941. At that point, the Trading with the Enemy Act diverted all payments to U.S. government.

The purportedly all-American press published the world’s most dangerous fascist, paying him as long as it was legal, but routinely turned a cold shoulder to those who would have been his victims, especially African Americans and Jews.

On an average day — this was before literary agents became unavoidable — all incoming submissions numbered between 10 and 25. Each was registered on an editorial blank, recording the dates, reader reviews or in-house deliberations and the verdict. The original manuscript was to be returned to sender, discarded or, in one or two lucky cases each month, sent back for revisions.

Arriving from anywhere between Kathmandu and Kansas, many submissions had little literary or edifying merit. “Terrible tripe!” a bored editor would scribble. At times, a quick tombstone doodle got added as a garnish.

But a surprising amount of “tripe” was publishable. The war-weary public, the editors anticipated, craved entertainment. Lugubrious pulp novels about spies and star-crossed lovers trapped in the sweltering colonial heat were deemed a “good sale.” As was “respectable, untalented historical fiction” with a “hollow ring [and] sleazy touch.” And so were memoirs of reputable war correspondents, despite being “commonplace” in the editors’ estimation.

Not all that was rejected was “tripe.” Vladimir Nabokov, an editor complained, “has never had but one plot.” Halldór Laxness’s sin was “Scandinavian melancholy de luxe.” Still, those white rising stars, even when rejected, were never reduced to types.

African American and Jewish authors, by contrast, were. Editors regularly dismissed black authors for sounding “the Negro lament.” Without a doubt, the black South African writer Peter Abrahams could write, but the editor could not “see anybody putting down $2.50 for his re-iteration of the negro lament, no matter how deeply the author feels what he is saying.” “As he is snubbed by the whites,” the editor went on, “he broods bitterly and sometimes poetically on the subject.”

Abrahams was not alone. Clear-eyed accounts about African American life by white authors were filed under the same rubric. A Smith College-educated social worker’s “drab and tragic story of Negro life,” a verdict went, was fated for “commercial failure.”

The same held true for “painfully Jewish” works produced by several generations of refugees. Among those who had arrived before 1933, Alexander Godin was pegged as “a profoundly serious Jew with bitter memories,” “monotonously tragic and so completely unrelieved by anything humorous.” His peer Florence Lipkin was derided for her “typical Jewish-Socialist novel.” Ruth Levine’s fictionalized pogrom reminiscence, “Machine-Gun Lullaby,” appeared “colorful” and “sharply drawn,” but in the end, “quite tedious.” “The scenes of the Jewish family,” the rejection snubbed, “are much better as they sit in darkness waiting for the looters to break down the shutters.”

Among the Holocaust memoirs, a 15-year-old French Jewish boy’s account of his family’s flight across the Pyrenees was hailed as “extraordinary,” with shattering details of a concentration camp and a grandfather’s suicide. “It would be a great pity if this Human Document were lost,” an editor noted before concluding, cruelly, that the book “doesn’t gain any sales appeal from the age of the writer; you don’t think about his age.” Lost it would be.

Trying to capture rejected authorship by race, ethnicity, gender, age or class, is a challenge. The records are scattered. Some are uncataloged or incomplete for reasons that not even archivists can explain. Some holdings are listed only on paper. The frequently used records of the published white men — such as Winston S. Churchill, in Houghton Mifflin’s case — have greater stature and higher priority.

As in the case of National Geographic and the New York Times, righting the book industry’s discriminatory wrongs will take more than piecemeal apologies on behalf of individual presses. The Association of American Publishers ought to insist on across-the-board scrutiny of the big publishers’ pasts. It must establish that sensitivity readers, hired to identify offensive cliches, work to respond to the racist legacy and are not “diversity hires” or, worse, censors.

To recapture the voices of the unpublished (and integrate them into the history of American literature), publishers must fund the study and digitization of corporate records via the archives and foundations that would safeguard the researchers’ independence. Archives must collaborate on collating the records into more accessible and searchable formats. A digital database capturing the rejected authors’ names and information through optical character recognition software is an important long-term goal. Posthumous publication of at least some minority writers rejected without a reason would be desirable whenever possible.

In addition, white readers must give their book-buying habits a cold, hard look. After all, it is this audience’s tastes that the editors struggled to anticipate. We, too, are part of the problem.