Historical research is challenging: you cannot simply go and see what happened, you have to trace back through a welter of documentary sources that may themselves have been selected in ways that impede your search for knowledge and understanding, that can be contradictory, and that are always only part of the potential evidence you might wish to have had.
That lesson, which is hardly news to practicing historians, historical archaeologists, or historical anthropologists, has been beautifully illustrated by newly published work by Rebecca Fried– a high school student. Quoting from the abstract of her article in the Journal of Social History:
Richard Jensen has forcefully argued that the absence of evidence supporting the Irish-American community’s historical memory of “no Irish need apply” restrictions in advertisements and signs suggests that these “NINA” publications, and particularly those directed to men as opposed to female domestics, did not occur to any appreciable extent in American history. Jensen argues that the NINA memory requires explanation as a psychological phenomenon rather than a historical one. This article surveys additional evidence from the nineteenth and early twentieth centuries documenting the publication of NINA-restricted solicitations directed to men. It shows that there were many such advertisements and signs… the earlier view of historians generally accepting the widespread reality of the NINA phenomenon is better supported by the currently available evidence.
I will admit that I was unaware that a historian had painted the Irish American memory of discrimination as a “psychological phenomenon.” When I went to read Jensen’s article — published in 2002 in the same journal — I was horrified. The article opens with the statement that “the Irish American community harbors a deeply held belief” that he ends by calling “an urban legend.” He describes Irish Americans in the 19th-century U.S. as having “a preindustrial non-individualistic group-oriented work culture” that appears to be the real focus of his scorn, calling Irish Americans “premodern”, and describing them as having a “‘chip on the shoulder’ mentality.” He engages in remarkably broad negative stereotypes, suggesting that the lack of upward social mobility documented for 19th-century Irish American immigrants was due to “internal self-defeating factors, such as heavy alcoholism, weak motivation, poor work habits, and disorganized family life.” After printing the lyrics to a popular song from 1862, he concludes “after a few rounds of singing and drinking, you could easily read the sign”.
But while this (and other) evidence of Jensen’s bias is worth noting (and others have done so), I am more interested in the way that his article envisions the utility of the written record, fails to consider the partial and biased nature of archival sources, and discounts the powerful and indispensable testimony of personal memory. At one point early in his article, he writes
It is possible that handwritten NINA signs regarding maids did appear in a few American windows, though no one ever reported one.
“No one ever reported one”: except for all those delusional Irish Americans who kept that memory alive. “No one” here does not mean no person: it means no documentary source that Jensen was able to identify. His own article names two individuals who said they personally had seen or been told of such signs: a former Speaker of the House of Representatives, Tip O’Neil, and former Senator, Edward Kennedy. Rather than credit their memories of early 20th century use of the formula, “No Irish Need Apply”, Jensen wrote that
physical NINA signs could have flourished only in intensely anti-Catholic or anti-Irish eras, especially the 1830—1870 period. Thus reports of sightings in the 1920s or 1930s suggest the myth had become so deeply rooted in Irish-American folk mythology that it was impervious to evidence.
Jensen doesn’t explain why he can assert that discriminatory signs could only have “flourished” at certain times. By using the terms “myth” and “mythology” instead of “oral tradition” or even “memory”, he adopts a position of judge of the veracity of memory and oral tradition, underlined by his closing claim that such memories were “impervious to evidence”. That phrase dismisses personal memory as “evidence” in favor of documents, understood in a far too simple way from the perspective of contemporary scholarship in history.
The fundamental mistake here is ignoring what are recognized as “silences in the archives“. In my discipline, one of the touchstones of the demonstration that textual sources cannot be presumed to be either complete or unedited came from Michel-Rolph Trouillet’s book Silencing the Past: Power and the Production of History. Historian Steven Lubar provided an extended look at Trouillet’s contribution and the continuing power of his analysis after Trouillet died in 2012. The book is widely recognized as providing a way to talk about how archives and historians, together, create narratives by silencing other possible stories.
As Kenneth Maxwell, reviewing the book in Foreign Affairs wrote in 1996, Trouillet showed how “power silences certain voices from history”. Writing in Social History, Robert Gregg described Trouillet as showing that “where there is silence, there must also be silencing”:
We cannot uncover all silences, and the choices we make as to what to uncover and what to leave in silence are political (they actively contribute to silencing the past).
That is a statement of responsibility that every historical researcher has, because the process of stitching together fragmentary sources requires assessment of what was left out, and is itself a part of the process of selection, editing, and selective sharing of “evidence”.
What Rebecca Fried did was start with the assumption that a widely shared oral tradition might reflect something that actually happened. Like Jensen, she investigated a number of places that the “No Irish Need Apply” phrase might have ended up being documented. Signs, after all, would only be discovered if someone thought to keep them and either deposit them in museums as historical artifacts (not a practice common for ephemeral artifacts of everyday life until the late 20th century) or passed them down as property (again, hardly to be expected). Photographs of such signs might be found incidentally in street scenes, but the task of winnowing through many historic photos would be challenging, and photographic records of street scenes are also, obviously, selective.
So Fried went to the same kind of documentary records as Jensen had: want ads in newspapers. But her search turned up many more instances of the phrase:
We have more NINA advertisements from the 1840s than from any other decade, but from the 1850s through the first decade of the twentieth century, the frequency of NINA-restricted advertisements remains generally similar.
Coverage of Fried’s research on the website Irish Central initiated a series of exchanges in the comments section in which Jensen responded, reiterating his claims and dismissing Fried’s findings as “a couple of drops of water at the bottom” of a “very big glass”.
Since his article is openly posted, while hers is behind a paywall, his assertions could have stood as refutation– except that Ms. Fried responded; always professionally, including this very sensible argument about why expecting anyone to find documentary evidence of the actual signs is problematic:
the vast majority of such signs would not be documented by newspapers at all for reasons discussed in the article. The surprise is that there are so many surviving examples of ephemeral postings rather than so few….I think the ordinary inference finding lots of signs and lots more newspaper advertisements drawn from sources that are demonstrably far from complete is that they look more or less as one would expect if the NINA phenomenon was real and sometimes pervasive. After such a showing, I do think that the burden should fall on you to show that mass delusion rather than ordinary memory should be invoked to account for this memory. I think the evidence, which includes many advertisements and signs, many more female-directed ones not even collected, and strong, concrete reasons showing that the existing digitized databases are vastly under-inclusive, strongly supports the ordinary, simple explanation rather than the unusual psychological explanation.
Fried sought and found not only newspaper ads, but print works mentioning seeing signs posted (the only way we should really expect such ephemera to enter the documentary record), news stories about labor actions in reaction to “No Irish Need Apply”, court cases stemming from such ads, and even Confederate propaganda seeking to persuade Irish Americans to support the South in the Civil War.
Fried’s article was published through the same peer-review process, and in the same journal, as Jensen’s earlier work. That means it met the standards of contemporary scholarship. When Jensen wrote his paper in 2002, it was plausible. The selection of documents promoted a particular explanation, which is always the case in historical writing. Fried showed that there are other documents that support a different story. Both historians labored under the reality that only a fraction of what happens is reflected in writing, only a fraction of what is written ends up in print, only a fraction of what ends up in print is preserved, and only a fraction of what is preserved is accessible to any scholar at a given time.
What makes the difference between their two interpretations of the documents they found is not in the documents themselves. One scholar, for presumably good reasons, believed that Irish Americans had promoted a self-exculpatory narrative of discrimination that never happened, and discounted oral tradition. The other began her research assuming that historical memory, while it may collapse time (so that events in a more distant past are kept alive as if they were more recent) and may recast stories from one place or person to another, nonetheless is not manufactured out of whole cloth.
I find Fried’s argument more convincing because her implicit model of human behavior is more believable. I cannot quite grasp how Jensen thinks an entire ethnic group successfully promoted a false memory, through the power of a popular song (aided by apparently widespread inebriation), or– the alternative narrative– as a cynical and calculated manipulation intended to promote political power. Does he imagine there were meetings where someone plotted this out?
For Fried to be right, we simply need to imagine that in a network of immigrants, the experience of a few people in many communities spread as personal testimony to discrimination that actually happened– not necessarily everywhere, not necessarily all the time, but certainly as a documented part of political discourse in the 19th century by groups like the Know-Nothing Party.
The formation of social memory would arguably have drawn power from the many examples of “No Irish Need Apply” used in print ads. As Fried writes, we should assume that what historians have found in the way of such ads is only part of what was printed: “existing digitized databases are vastly under-inclusive”. One blogger writing about the new research looked for similar ads in the online archive of a newspaper not included in either Jensen’s or Fried’s articles, and found them there as well. There is no reason to assume that the sentiment was only acceptable in print ads and did not find its way into placards used at sites of employment.
Experiences by some Irish Americans with print ads, signs, and other forms in which anti-Irish sentiment circulated would have resonated with others in the community. These experiences are reflected in the material archive indirectly, through articles about legal cases, in fiction, and yes, in popular song. But in a world in which “No Irish Need Apply” was acceptable language in newspaper ads, it wouldn’t take a mass delusion to foster a memory of discrimination.
Demanding one form of evidence of that discrimination in a world in which much of the material of everyday life simply rots away is, to paraphrase Trouillet, actively contributing to silencing groups and individuals who have unequal access to the means for the production of historical narratives. By all means, interrogate historical memory; but do so with a sense of the power of personal narration to keep alive things that may never make it into the official transcripts, or only be reflected there in muted ways.