Just came across news of the publication of Rebecca Fried’s excellent article “No Irish Need Deny: Evidence for the Historicity
of NINA Restrictions in Advertisements and Signs”, Journal of Social History, 10:1093, 2015, from @seth_denbo on Twitter.
First, the background to this article. Fried’s essay is a refutation of a 2002 article by the historian Richard Jensen that claimed that “No Irish Need Apply” signs were rare to nonexistent in 19th Century America, that Irish-American collective memory of such signs (and the employment discrimination they documented) was largely an invented tradition tied to more recent ideological and intersubjective needs, and that the Know-Nothings were not really nativists who advocated employment (and other) discrimination against Irish (or other) immigrants.
Fried is a high school student at Sidwell Friends. And her essay is just as comprehensive a refutation of Jensen’s original as you could ever hope to see. History may be subject to a much wider range of interpretation than physics, but sometimes claims about the past can be as subject to indisputable falsification.
So my thoughts on Fried’s article.
1) Dear Rebecca Fried: PLEASE APPLY TO SWARTHMORE.
2) This does really raise questions, yet again, about peer review. 2003 and 2015 are different kinds of research environments, I concede. Checking Jensen’s arguments then would have required much more work of a peer reviewer than more recently, but I feel as if someone should have been able to buck the contrarian force of Jensen’s essay and poked around a bit to see if the starkness of his arguments held up against the evidence.
3) Whether as a peer reviewer or scholar in the field, I think two conceptual red flags in Jensen’s essay would have made me wary on first encounter. The first is the relative instrumentalism of his reading of popular memory, subjectivity and identity politics. I feel as if most of the discipline has long since moved past relatively crude cries of “invented tradition” as a rebuke to more contemporary politics or expressions of identity to an assumption that if communities “remember” something about themselves, those beliefs are not arbitrary or based on nothing more than the exigencies of the recent past.
4) The second red flag, and the one that Fried targets very precisely and with great presence of mind in her exchanges with Jensen, is his understanding of what constitutes evidence of presence and the intensity of his claims about commonality. In the Long Island Wins column linked to above, Jensen is quoted as defending himself against Fried by moving the goalposts a bit from “there is no evidence of ‘No Irish Need Apply'” to “The signs were more rare than later Irish-Americans believed they were”. The second claim is the more typical sort of qualified scholarly interpretation that most academic historians offer–easy to modify on further evidence, and even possible to concede in the face of further research. But when you stake yourself on “there was nothing or almost nothing of this kind”, that’s a claim that is only going to hold up if you’ve looked at almost everything.
I often tell students who are preparing grant proposals to never ever claim that there is “no scholarship” on a particular subject, or that there are “no attempts” to address a particular policy issue in a particular community or country. They’re almost certainly wrong when they claim it, and at this point in time, it takes only a casual attempt by an evaluator to prove that they’re wrong.
But it’s not just that Jensen is making what amounts to an extraordinary claim of absence, it is that his understanding of what presence would mean or not mean, and the crudity of his attempt to quantify presence, that is an issue. There may be many sentiments in circulation in a given cultural moment that leave few formal textual or material signs for historians to find later on. Perhaps I’m more sensitive to this methodological point because my primary field is modern Africa, where the relative absence of how Africans thought, felt and practiced from colonial archives is so much of a given that everyone in that field knows to not overread what is in the archive and not overread what is not in the archive. But I can only excuse Jensen so far on this point, given how many Americanists are subtle and sensitive in their readings of archives. Meaning, that even if Jensen had been right that “No Irish Need Apply” signs (in ads, in doors, or wherever) were very rare, a later collective memory that they were common might simply have been a transposition of things commonly said or even done into something more compressed and concrete. Histories of racism and discrimination are often histories of “things not seen”.
But of course as Fried demonstrates comprehensively, that’s not the case here: the signage and the sentiment were in fact common at a particular moment in American history. Jensen’s rear-guard defense that an Irish immigrant male might only see such a sentiment once or twice a year isn’t just wrong, it really raises questions about his understanding of what an argument about “commonality” in any field of history should entail. As Fried beautifully says in her response, “The surprise is that there are so many surviving examples of ephemeral postings rather than so few”. She understands what he doesn’t: that what you find in an archive, any archive, is only a subset of what was once seen and read and said, a sample. A comparison might be to how you do population surveys of organisms in a particular area. You sample from smaller areas and multiply up. If even a small number of ads with “No Irish Need Apply” were in newspapers in a particular decade, the normal assumption for a historian would be that the sentiment was found in many other contexts, some of which leave no archival trace. To argue otherwise–that the sentiment was unique to particular newspapers in highly particular contexts–is also an extraordinary argument requiring very careful attention to the history of print culture, to the history of popular expression, to the history of cultural circulation, and so on.
Short version: commonality arguments are hard and need to be approached with care. They’re much harder when they’re made as arguments about rarity or absence.
5) I think this whole exchange is on one hand tremendously encouraging as a case of how historical scholarship really can have a progressive tendency, to get closer to the truth over time–and it’s encouraging that our structures of participation in scholarship remain porous enough that a confident and intelligent 9th grader can participate in the achievement of that progress as an equal.
On the other hand, it shows why we all have to think really carefully about professional standards if we want to maintain any status at all for scholarly expertise in a crowdsourced world. I’ve said before that contemporary scholars sometimes pine for the world before the Internet because they felt safe that any mistakes they make in their scholarship would have limited impact. If your work was only read by the fifty or so specialists in your own field, and over a period of twenty or thirty years was slowly modified, altered or overturned, that was a stately and respectable sort of process and it limited the harm (if also the benefit) of any bolder or more striking claims you might make. But Jensen’s 2002 article has been cited and used heavily by online sources, most persistently in debates at Snopes.com, but also at sites like History Myths Debunked.
For all the negativity directed at academia in contemporary public debate, some surveys still show that the public at large trusts and admires professors. That’s an important asset in our lives and we have serious collective interest in preserving it. This is the flip side of academic freedom: it really does require some kind of responsibility, much as that requirement has been subject to abuse by unscrupulous administrations in the last two years or so. We do need to think about how our work circulates and how it invites use, and we do need to be consistently better than “the crowd” when we are making strong claims based on research that we supposedly used our professional craft to pursue. It’s good that our craft is sufficiently transparent and transferrable that an exceptional and intelligent young person can use it better than a professional of long standing. That happens in science, in mathematics, and other disciplines. It’s maybe not so good that for more than ten years, Jensen’s original claims were cited confidently as the last word of an authenticated expert by people who relied on that expertise.
Commonality arguments: see also Time on the Cross.
Yes, good example.
Several comments:
1. Rebecca Fried: WOW. Her research is better, and so is her prose.
2. Mark Twain used NINA as part of a joke in 1871 (Roughing It, Vol II,
Chapter VI), good evidence that it was a commonly-known phrase at that time.
3. I’m not sure what “commonality” means, but if it is “frequency of
occurrence” then yes, historians have a problem thinking about it–like
most people, who don’t think statistically. Another example:
the issue of “how big is a manor?”: see J. H. Hexter’s article
“Storm over the Gentry”. (NB I’m in an STEM area that is heavily
statistical).
4. The ads are fascinating in their preference for “German Protestant or
colored” over Irish Catholic. I sense a possible PhD thesis and tenure
book here; see (1) above.
5. I can think of plenty of dubious claims in my own field that survived
untested for 10 years or more–though usually because they aren’t that
generally interesting.
You know, I want to yell in frustration, “How did this even make it past the referees?!?” but I sort of assume the referee was tired, had a ton of other stuff on his/her plate, and figured that if he said he’d been through all the material, then there was no need to check up on it.
Ms Fried did a very impressive amount of google searching. But she is very weak on historical context and got it wrong. The NINA myth I examined is that “Unskilled workers and servants, especially, encountered the ubiquitous ‘No Irish Need Apply’ notices when they searched for jobs in Boston, New York, and other major cities.” (quoting Kerby A. Miller, Emigrants and Exiles (1985), p 323). I showed the NINA slogan was widespread in the US, Britain and Canada, and was often used in a jocular way. She found 69 examples from 21 cities. I looked at each one and conclude that she found exactly one actual help-wanted window sign in the 19th century. (She also has an inside bulletin board sign in 1932 in her case #69 from Glens Falls NY.) It appeared in 1883 in Port Washington in rural Ohio, where the local police department advertised a patronage job to clean lamps on bridges. In those days the patronage system distributed jobs to the winners of local elections. When Republicans won, they replaced all the Democrats, and practically all the Irish were Democrats, so that “no Irish need apply” was the appropriate slogan after a local election won by Republicans. This Ohio case is only actual 19th-century sign that Fried has identified, and until now it was quite unknown to Irish outside Tuscarawas County, Ohio.
Most of her examples are classified ads from the help wanted pages of online newspapers. A third of her cases come before 1845 when the Irish started arriving, and she and I both think they may be the product of two local employment agencies. By my count, Fried found 21 ads for actual jobs in American cities, 1845-1932, not counting farms and private household jobs. She has four ads from Boston, three in New York City, two in Chicago and St. Louis, and one each in Cincinnati Cleveland Philadelphia and Washington–that’s it in an 87 year stretch. For all the 19th century newspapers that are online, I calculate one would have to read on average 41,000 pages of small print to see even one little classified ad. I call that very rare, but to Fried they were “often common.” It’s Miller who went on a limb in 1985 by calling NINA signs “ubiquitous” in major cities when he never saw one. Furthermore, I doubt Miller ever saw a NINA newspaper ad for men.
What I wrote in 2002 regarding window signs was:
“The complete absence of evidence suggests that probably no such signs ever existed at commercial establishments, shops, factories, stores, hotels, railroads, union halls, hiring halls, personnel offices, labor recruiters etc. anywhere in America.” [p 405]
I DID qualify the statement as “The complete absence of evidence suggests.” I am willing to change my mind when evidence appears. In my judgment there is still a complete absence of evidence. (Note that the myth is about signs not newspaper ads; I was the first to find a newspaper ad for men—I said those were rare but did exist and that NINA ads for women were much more common.)
Professor Burke seems to believe Fried found a genuine help-wanted NINA window sign from the 19th century. He says: “Fried demonstrates comprehensively, that’s not the case here: the signage and the sentiment were in fact common at a particular moment in American history.” Then he berates me at length. Ok—I think he took that from the misleading press release. Did he even glance over Fried’s 69 cases?? I challenge him to tell us all exactly which window sign in Fried’s list he thinks is genuine. If he cannot do so his long attack on me is would appear to be highly unprofessional.
Richard Jensen
Ps. Professor Burke praises Rebecca Fried the “high school student at Sidwell Friends,” saying “a confident and intelligent 9th grader can participate in the achievement of that progress as an equal.” She is indeed impressive. However she wrote the essay while a Middle School student in the eighth grade; she will enter the 9th grade a month from today. Does this blunder demonstrate the usual research skills of a Swarthmore professor?
Sorry for having comments stuck in queue while I was travelling.
Re: Professor Jensen’s comments. I think “tendentious” is perhaps a fairly good adjective for the spirit in which they’re offered, and that would be one reason why I don’t think his interpretation of the history he’s examining holds up very well. He moves his evidentiary goalposts a great deal–at one moment a very minute difference-that-makes-no-difference is sufficient for him to passing sweeping judgment, at another moment a small difference is easily dismissed by him because it doesn’t fit the sweeping judgment he wishes to pass.
I want to return somewhat insistently to the larger methodological problem I raise in my own response, because it applies to more than just Professor Jensen and his essay. Namely, what do we as historians accept as evidence of presence when we’re talking about ephemeral material culture or equally ephemeral and/or subcultural forms of speech and everyday practice?
As I observed, Africanists have been forced to think very hard about this problem. Most of the archival sources we have to work with actively deny the humanity of the principle subjects of our research, and in any event take little interest in the everyday discourses, practices and experiences of those subjects. But we do not conclude from this absence in the archives that there is absence in historical reality–we do not conclude that Africans had no everyday practices, lives or discourses. At the same time, you can’t just make it up from nothing. But sometimes just a few references in existing material are enough to open up a line of interpretation, to catalyze oral history or ethnography, to suggest new kinds of documents to look at, to see material culture in a new light. More importantly, sometimes just a few references are sufficient to plausibly argue that a particular way of speaking, a particular kind of practice, a particular sort of artifact, was in fact common despite the fact that it’s barely mentioned or documented in the evidence we have to work with.
Stephanie Newell and Osumaka Likaka, for example, have written persuasively about colonial-era practices of naming, about how Africans subversively renamed and spoke about imperial officials and activities. Like anything that’s a “weapon of the weak”, this kind of naming barely registered in the awareness of the imperial archive–or was misread. But that misreading or minimal recognition is good enough to suggest a hidden ubiquity, I think.
So to the question of signs, before we get to ads. Professor Jensen seems to me to have an expectation about commonality that he discusses very little in his original piece, that has if anything deepened in his responses to Fried (and here to me as well). The signs, he believes, either should have survived in some existing collection of antebellum material culture or should have been commonly remarked upon by eyewitnesses in many instances. Otherwise, he argues, it is reasonable to argue that they didn’t exist. But this seems to me to move far too fast past a really difficult problem in historical interpretation. First, that some forms of material culture that we are fairly sure existed simply do not survive; second, that some forms of material culture and everyday practice are not commonly remarked upon by eyewitnesses in various kinds of texts even though they are fairly ubiquitious. Indeed, in some cases *because* they are ubiquitious.
But also in some cases because they are primarily noticed or experienced by subordinate or marginalized subjects. In the case of NINA signs, both problems are in play. For example, how many ‘help wanted’ signs of any kind do we have in existing collections from antebellum America? How often was employment offered via signs in windows at all? How often was employment offered through conversations or direct approaches in relationship to signs? Could some of the memory of “signs” derive from everyday conversation? How much do we know about the denominators in relationship to all the numerators that seem so significant to Professor Jensen? (And to Ms. Fried?)
I drove through several small New England towns in the last week and saw many handwritten “help wanted” signs in the windows of businesses. But were I not having this conversation, it would never occur to me to record or comment upon having seen them. None of those signs will survive for long–I suspect most will be in the garbage within several months. They’re informal, quickly composed, etc. If a century from now it was concluded that American businesses primarily filled their vacancies using Craigslist and Monster.Com or perhaps the want ads in newspapers, that would be in some sense very wrong when it came to small businesses in smaller communities.
The problem of subordinate or marginalized experience is more potent here. Professor Jensen insists with great assertiveness on extensive “evidence of presence” before he will countenance the possibility that signs existed, but dispenses with the need for direct evidence of the actual experience of Irish-Americans between 1840 and 1862 quite handily. Meaning, he is willing to argue that if he can establish evidence of the absence of signs, he is quite free to make strong claims about the process by which collective memory was established among Irish immigrants through indirect evidence–that there was a song by 1862 which had an antecedent form, that immigrants *had* encountered the signs in England, and that there are general sociological claims that one may make about “the invention of tradition”. But if there is no court case of an insurgent Irish youth throwing a brick through an offensive window, neither is there a transcript of Irish immigrants engaged in the construction of memory. Which sounds as if it’s a tall demand, but in fact the historiography of memorialization and the “invention of tradition” often finds surprisingly direct evidence of the conversations, discourses and relations by which people come to believe in shared experience.
I think it’s important in general to start from the premise that if people at a certain moment in the past seem to believe very strongly that something happened to them, it may well have happened in some form fairly close to the text of that memory. It is especially important to do that for dominated, subordinated or marginalized communities. How often has dominant opinion judged the reported experiences of such communities to be fictional or exaggerated only to be compelled at a later date to recognize that what was attested to was in fact quite common?
It’s also very rare for memories to be entirely functional, ideological or just the epiphenomenal traces of some overall process of social formation–their content comes from something real. Now Professor Jensen argues that there is a “real” referent here–the memory, both direct and reported, of experiences in England and possibly Canada as well, or for male workers, the “real” referent of the experiences of Irish-American women seeking domestic work. But he’s very quick to conclude that otherwise the memory is a fiction and a very interested fiction at that.
We are right now in the middle of considerable political and social upheaval in the United States over the fact that long-reported experiences of African-Americans with police violence are being documented in a new way, through ubiquitious video, which I think should give anyone pause about where to place their trust–in the reported experiences and memories of people whose perspectives do not compose the archive to a great degree, or in what the archive says about them.
This is also why I think even a few cases of seen, reported or described “real” signs in Fried’s article matter so much, and why it’s important not to create a benchmark for ubiquity that’s both unrealistic and unfair. How often in Professor Jensen’s measurement would an Irish-American man in the 1840s or 1850s need to have had an experience in which he directly was told “no Irish need apply” (in that phrasing or closely related ones) for him to conclude this was a common explanation for his circumstances? Once in a lifetime? Once a month? Is that a general benchmark for documenting discrimination in history? Would some subordinate or marginalized people be perfectly right to conclude that they are subject to a discriminatory practice even if the only direct evidence is something that happened to someone else in their community? Do you have to wait to be personally oppressed to ‘remember’ that your community is oppressed? If, for example, there was an employment agency in the early 1840s using the NINA formulation in ads fairly commonly, why wouldn’t that be a legitimate basis for a larger, later group of immigrants “remembering” that they might be subject to discrimination? The ads might become more subtle–and Fried documents that there were also significant objections to them, something that the 2002 article says it expects to find but does not, so the objections might also have driven the ads underground or into less-explicit forms. But the early ads would still serve as a credible basis for knowing and “remembering” that there was some discrimination.
After all, discrimination is to some extent by its nature a practice of something not happening, of jobs not offered, housing refused, rights not accorded. And mostly in our own times–and I think others–the everyday practice of discrimination goes on without great ideological fanfare and without a blazingly clear archival trail. That is precisely what has made it so hard in our present day to contest employment and housing discrimination within the legal system, and thus why many who are discriminated against simply accept it as a social fact. A given landlord can refuse to rent to African-Americans who make inquiry and as long as they make no statement to that effect, each individual who is turned down is left without an overall sense of the pattern. So when there are moments of rupture or contradiction–when the system does provide a sign or an ad or a statement–I think it’s often fair to see those moments as the visible tip of something pervasive but invisible.
Also a small matter but seriously, characterizing Fried’s research as “an impressive amount of google searching” does you no credit on several levels. If you want a yardstick for being unprofessional, use that as a good measure to start with. It’s not only ungracious and inaccurate, but it rebounds on your own methodology in the 2002 piece in a rather serious way.
would appear to be highly unprofessional.
I do love the conjunction of the above with the PS, below:
Ps. Professor Burke praises Rebecca Fried the “high school student at Sidwell Friends,” saying “a confident and intelligent 9th grader can participate in the achievement of that progress as an equal.” She is indeed impressive. However she wrote the essay while a Middle School student in the eighth grade; she will enter the 9th grade a month from today. Does this blunder demonstrate the usual research skills of a Swarthmore professor?
“You’re unprofessional and you got her grade wrong by one year! Nyah Nyah Nyah!”
Okay, then.
Via Miriam Elizabeth Burnstein, I see that the scholar Liam Hogan is adding some very nice complexity and texture to this debate. See his Twitter feed, https://twitter.com/Limerick1914.
Timothy Burke writes: “It’s also very rare for memories to be entirely functional, ideological or just the epiphenomenal traces of some overall process of social formation–their content comes from something real.”
While Timothy grounds his argument on African research, it is a very important point in the history of East Asia as well. Tessa Morris-Suzuki’s recent Japan Focus article “You Don’t Want to Know About the Girls? The ‘Comfort Women’, the Japanese Military and Allied Forces in the Asia-Pacific War” makes this point well:
“The question of historical sources lies at the heart of the bitter contemporary disputes about the “comfort women”. Those who deny that the Japanese military forcibly recruited women to brothels, or who dismiss the issue out of hand, commonly draw a sharp distinction between official documents and oral testimony, discounting the latter or focusing their energies on parsing specific pieces of testimony in order to undermine their credibility. But both written and oral historical source materials are created by human individuals with human strengths and weaknesses. Why should we privilege one over the other? The bureaucrats who compose an official document are neither more nor less likely to present a full and accurate picture of the truth than elderly women who tell the stories of their past.”
http://www.japanfocus.org/-Tessa-Morris_Suzuki/4352/article.html
An inquisitive mind and honesty are worth more then a doctorate. Too bad we make so little effort to search such people out.
This cuts to the core of the methodological problem in Jensen’s argument:
“First, that some forms of material culture that we are fairly sure existed simply do not survive; second, that some forms of material culture and everyday practice are not commonly remarked upon by eyewitnesses in various kinds of texts even though they are fairly ubiquitious. Indeed, in some cases *because* they are ubiquitious.”
In short, he is conflating an absence of evidence for evidence of absence.
Jensen’s inquiry would be better served by asking the question of why NINA signs are not abundant in historical photographs or archives. Plausible explanations actually exist for each. Photography was very rare prior to the 1860s, and even then almost always used for portraiture – not randomly capturing “help wanted” signs in the windows of buildings. As to archival collections, I’d suggest that it was no more common to deposit an old handwritten paper NINA sign at the local library in 1860 than it would be for you to do the same today with a cardboard sign following the conclusion of a yard sale.
The far more interesting evidentiary attestation is, of course, the extended list of NINA newspaper ads that Fried has uncovered. Any decent statistician would note that, in addition to being records in their own right, NINA ads (due to their high survival rate) are also likely a proxy measurement for other types of discriminatory employment practices that were less likely to be recorded for the aforementioned reasons. There is no reason to believe that an employer who posted a NINA ad in a newspaper would withhold such stipulations from a paper sign hung in his window, or a verbal disclaimer provided in a job interview. In that sense, the NINA ads are themselves an attestation to the prevalence of sentiments that likely appeared in other less reliably recorded or retained forms.
Thanks, Phillip. I’m also honestly curious about whether there’s a good understanding of how job availability was communicated and offered in general between 1830-1880 in the U.S., at a time of enormous change generally in print culture, in commercial and industrial life, and in the layout of urban space. The stubborn literalness of Jensen’s obsession with the sign-in-the-window is here not helping much at all with the question of whether Irish collective memory of responses to seeking employment is based on some set of real experiences. Over time, it’s not uncommon for a disparate set of real memories to ‘migrate’ to a single, concrete focal signifier (here a word with entertaining double meaning in this conversation), and even if one can later demonstrate that the symbolic container of those memories is exaggerated, it doesn’t mean the memories are false. It’s a bit like screaming “ROSEBUD IS JUST HIS SLED” and thinking that you’ve just exposed “Citizen Kane” as narratively incoherent. If it turns out the sentiment was commonly expressed to Irish migrants seeking work, that there were significant numbers of want ads using the language, etc., then it seems to me that getting this obsessed with the literal sign-in-a-window is really missing the point, even given that I’m not convinced that Jensen’s faith in absence-of-evidence is in any sense evidence-of-absence, even if his faith in the former is at all reasonable.
I have nothing to say about the case in point, but I offer an anecdoete about the general problem. Some years ago, say a decade or so, I visited the laboratory of Ralphy Halloway at Columbia. Ralph is a physical anthropologist and an expert in the evolutionary history of the human and proto-human brain. He told me that, at the time, his laboratory held about a tenth of the world’s collection of (relevant) hominid skull fragments and associated endocasts. I forget the number. He also told me that, as a percentage of the estimated population of hominids of which the worldwide collection is evidence, the collection is very very small, less than .01 percent. Nor do we have any reason to believe that these surviving fragments represent a random sample of the original population. That’s what we’re using to guestimate about the evolution of the human brain.