Wednesday, March 02, 2005


Emma Dunham Kelly-Hawkins

Scott McLemee's newest column at Inside Higher Ed discusses the recent discovery by a Brandeis graduate student (go, grad students, go!) that Emma Dunham Kelly-Hawkins, long presumed to be a turn-of-the-century African American novelist, may have been white. Other good posts on the subject can be found at Crooked Timber, Easily Distracted, and The Reading Experience. Ralph Luker also asks whether there is a history scandal here somewhere.

Most of what I would say about this news has been covered in these places, but there are a few tangents that I haven't seen raised elsewhere.

First (and maybe I'm sticking my neck out for saying this), there seems to have been an enormous rush to judgment that the evidence provided by Holly Jackson, the Brandeis graduate student aforementioned, is straightforward proof about Kelly-Hawkins' race. [P.S. The preceding was a sloppy sentence. See comments for clarification.] But it's interesting to me that almost all of the evidence for her "whiteness" hinges on nineteenth-century census records.

This is not the only evidence, of course: there are documentary records at the end of Kelly-Hawkins' life, along with family memories that identify the family as white. But this later evidence could be explained, conceivably, by the family's conscious adoption of "whiteness," a not uncommon attempt in American history to erase African American ancestry from the family tree. Jackson considers this possibility -- that the family was "passing" -- but rejects it on the slender hypothesis that they could not have fooled the census-takers in a small Massachusetts town, where Kelly-Hawkins' family had lived for more than one generation when she was born.

Before accepting this hypothesis [P.S. Another sloppy phrase; imagine reading the sentence without it. See comments for clarification.], I'd like to know more about the way the census was taken in Massachusetts at the time. In Maryland, for instance, my understanding is that the census was usually recorded in the antebellum period by hired census-takers, who went (more or less) from door to door, asking for names and ages. Presumably, they sometimes also asked for "race," since there was a column on the census for recording this, usually "W" for white, "B" for black, and "M" for "mulatto." But the column was usually labelled "color," not "race," and it's highly probable that white census-takers often simply identified a person's "color" with their own eyes. That is, if a person looked white, the census taker could mark down his "W" and move on, regardless of the person's own identification of himself or herself.

Again, I don't know whether this was the way the census was taken in Kelly-Hawkins' case, but it's a question worth raising. [P.S. The preceding sentence gets closer to the main point I wanted to make.] I also don't know whether census takers necessarily knew the locals, as Jackson seems to assume. But my main point here is to question how easily many people still seem to take the census as a final arbiter of vital statistics or racial identities, when the census itself was a document shot through with ambiguity and power relationships. (Martha Hodes at New York University had an excellent article on these subjects two years ago in the American Historical Review, "The Mercurial Nature and Abiding Power of Race." The link will only work if you have an individual or institutional subscription to the History Cooperative.)

Here I would echo Burke's post that "race" itself is a cultural construction, a marker of identity rather than a signifier of determinate reality. Maybe I'm going out on a limb again by saying that there may have been some value in the scholarship that has tried to deal with Kelly-Hawkins' literature on the assumption that she was a black woman. Most of that scholarship has wrestled with figuring out the "aggressive whiteness" (to use Jackson's phrase) of Kelly-Hawkins' characters. This is the kind of scholarship that has helped us shake free of the idea that "whiteness" is a simple fact that inheres essentially in a person. Rather, "whiteness" could potentially be "aggressive," which means that it is a fluid and shifting concept that can be adopted, to various degrees, by individuals and even groups. Race is, to use Hodes' word, "mercurial." It would be a shame if somehow these new documents about Kelly-Hawkins undid that valuable insight. It would be a step backwards, rather than a step forwards, to reify race again, and make Kelly-Hawkins' (or anyone's) race a simple matter of reading letters in a column.

Second, I agree with McLemee that this new evidence does not mean (as Henry Louis Gates seems to have suggested) that Kelly-Hawkins' mediocre novels are no longer worth studying. McLemee points out that these novels are still worth studying as embodiments of the "banality of evil" and racism at the end of the nineteenth century. I would add, however, that we should not simply assume that because an author was white, he or she had no influence on contemporary African American writers. That seems to me to be a larger problem with the project of constructing a vacuum-sealed canon of African American literature: it assumes that the only tradition that matters for understanding black writers is the tradition built by black writers. Jackson writes, for instance, that Gates originally introduced the series in which Kelly-Hawkins appeared by saying that he had found, in these obscure black writers, "the literary ancestors of Zora Neale Hurston, Alice Walker, and Toni Morrison."

That way of tracing literary ancestry seems to me misguided. Is there any evidence that Alice Walker read Kelly-Hawkins? Probably not. It seems at least more plausible that Kelly-Hawkins' black contemporaries, like Frances E. W. Harper, might have read her work. That would at least be a good question for research. To find out that an author was white does not immediately prove her work's irrelevance to the African American canon, because black authors read white authors, and vice versa. Dickson Bruce has argued, for instance, that the real question for scholars of African American literature is to "investigate the historical conditions for an African American literary enterprise," and that means figuring out the "process in which black and white writers collaborated in the creation of what I call an 'African American literary presence.'" Even if Harper had not read Kelly-Hawkins, her novels probably had to deal in complex ways with the portrayals of African Americans in them. More evidence would be needed to show that Kelly-Hawkins had any influence, whether direct in Gates' sense or indirect in Bruce's sense, on the development of the African American canon. To show that she was white does not automatically prove she had no influence at all.

Collective Improvisation:
The census was taken just as you described, in every state. The census forms were exactly the same throughout the country, and the census was taken within the same time frame in every location. The census from 1790 to 1840 were heads-of-household listings only, with columns for how many others of a certain age range also lived in the household. Depending on the year, there were also columns for slaves owned, and "free" people of any color other than white who lived in the household. From 1850 onward, each member of the family was enumerated, and there were no "special" columns; if there were workers in the household they were enumerated within the same household, after the family members, and with their relationship ("domestic" or "servant," etc) listed. There was a column called "color" from 1850 through 1880. In 1850 and 1860 there were three options: W(hite), B(lack), M(ulatto). Indians were often listed as mulatto, for lack of something better. Chinese people really threw them for a loop, until 1870 and 1880 when the options became W(hite), B(lack), M(ulatto), C(hinese) and I(ndian). From 1900 throgh 1930, the question was "Color or Race," but in all my census reading, I've never seen anything besides W, B, M, C, I in those slots.

I probably didn't read the article carefully enough, but I wish it would have stated her parents' names (which I'm sure I could find if I actually looked hard enough for it!) and grandparents' names, so I could just pull all the census records right now and see for myself. I don't know a darn thing about the woman, but I'm a census-reader and like mysteries. :) The article stated at one point that the column for race was blank, and tallied at the bottom -- the tally isn't something you'd see often in 1860 (it was more in the pre-1850s when the lines on the form were not equal to the number of people actually enumerated on the sheet) but leaving an entire column blank was done out of sheer means "everyone here is the same." I see that a lot with the "Birthplace" question -- if the censustaker's notes say that almost everyone on that street was born in Massachusetts, then when he went back to transcribe the notes onto the original he may very well have put "Ma" at the top and nothing else for the other 60 or so lines on the page...even if one of them was born in "Ct," he wouldn't want to screw up his enumeration. Same with leaving blank the race...if he remembered seeing everyone as white, leave it blank and assume everyone was...

Posted by JM

Posted by Anonymous Anonymous on 3/02/2005 11:02:00 AM : Permalink  

Thanks for all the information, Julie! All of this just goes to show how a seemingly straightforward record like the census could be extremely messy and variable over time and space. You'd probably like the Hodes article I mentioned. After finishing this comment I'm going to find the link to it, but unfortunately it will only be available through an institutional library subscription.

I was also a little confused by the "tally" comments in Jackson's article. If you do track down the census records, I'd be interested to know what you find. If the evidence for her "whiteness" is actually a blank space on the census, all the more reason to wonder whether the case can just be so easily closed. 

Posted by Caleb

Posted by Anonymous Anonymous on 3/02/2005 11:14:00 AM : Permalink  

Oh good lord, the census is so iffy it's not even funny...yet it's so utterly important and relied upon by so many researchers for so many different things, funny turns to sad! Since the "factual" information is given by whomever answers the door that day, you have people who are 24 years old in one census and 40 in the next, or born in Ireland in one census and Louisiana the next, etc.

In the post-1900 census returns, Emma Hawkins (and her children) each have a very clear W in their "color or race" columns. I haven't found her in the 1870 or 1880 census returns because "Massachusetts" is a little vague given the number of _Irish_ Kelly families there at that time, and I don't know her parents' names. But in Dennis, Barnstable Co., Mass in 1870, no one was enumerated as "colored",in a town of over 3300 people. That doesn't mean there weren't non-white people, it means that everyone was enumerated with a W or without a mark in the column (with W assumed). In all of Barnstable Co, there were only ~360 out of ~33000 enumerated folks who were labeled "colored", so...I don't know if there just weren't a lot of non-whites in the area in 1870 or if the census takers were lazy. In all of Mass in 1870, there were ~7850 folks w/ the last name Kelly/Kelley, only 12 were listed as "colored". Just going by those numbers, I'd think you'd have to find overwhelming evidence to the contrary that a Kelly in Massachusetts in 1870s wasn't white.

Of course, I'm completely talking out of my rear, here, and don't know what all the woman at Brandeis looked at, etc. Incidently, W and M can look very similar in 1850/1860 census returns. Some of those folks had horrible handwriting. 

Posted by JM

Posted by Anonymous Anonymous on 3/02/2005 04:08:00 PM : Permalink  

Wow, Julie, you've done the hard work of researching my post for me! :-)

I think the main point I want to make in the post is not that Jackson is definitely reading the census wrong, but that reading the census doesn't always produce clear right or wrong answers. This isn't to say it's a worthless historical source, by any means, but only that it needs to be approached cautiously and with all these relevant wrinkles in mind. And it seemed to me that the stories surrounding this case were just taking for granted the census data as a sure-fire to settle questions about race.

Incidentally, I wonder why, if "W" was sometimes assumed in a given census, the "W" was ever specified at all in that census. Could it be that in such censuses -- where there are blanks and "W"s -- those that are explicitly marked might have been persons that the enumerator felt the need to specify a designation for, perhaps because their identity was (to the enumerator) ambiguous. That's purely speculative, of course, and it's hard to know how we can get in the heads of these enumerators. But the fact that their heads mattered shows how wrinkly the census can be.

Thanks again for all the info! 

Posted by Caleb

Posted by Anonymous Anonymous on 3/02/2005 04:19:00 PM : Permalink  

I'll shut up here in a moment, but I should also have said "according to the indices" with regards to my numbers above. There are discrepancies between numbers shown in indices and the numbers on actual returns, etc. But the point you made is THE point to make WRT the census: approach with caution. Lots and lots of caution.  

Posted by JM

Posted by Anonymous Anonymous on 3/02/2005 04:31:00 PM : Permalink  

Very thoughtful post, Caleb, and good discussion. I agree with you and Julie that a) it is a mistake to reify race; and b) that evidence from census records alone is not despository. On the other hand, we are referencing a society in which race was reified and Jackson is more familiar with the range of evidence about Kelley-Hawkins than any of us, so far, at least. What bothers me about Caleb's argument is that, apart from an author's reference in 1955, there is simply _no_ reason to think that Kelley-Hawkins might have been anything other than white. Why, then, hang onto that possibility? She makes no claim to an Afro-identity, there is no Afro-race consciousness in either of her novels, her family believes itself to be white, all known primary sources indicate that she and her family were white. Given all that, isn't it just hanging on to a scholar's blunder to say that the question remains open? By the way, I don't _lightly_ suggest that Henry Louis Gates does blindered research. You can be certain that he had research assistants, funds, and capabilities that would be the envy of any of us. With respect to Kelley-Hawkins, why wouldn't the content of the novels have raised elemental questions about the ethnicity of their author? With respect to The Bondwoman's Narrative, despite all the research resources and having turned up no convincing evidence, Gates plunged ahead with the publication-claims that he was predisposed to believe. Why bother with research at all, if you're going to do that? Simply _because_ it _might_ have vindicated your predispositions? That's the sort of post-modern madness that rightly drives non-academics nuts. And then we call them nuts. 

Posted by Ralph Luker

Posted by Anonymous Anonymous on 3/02/2005 06:06:00 PM : Permalink  

Ralph, I completely agree that Jackson is much more qualified to speak on these matters than I am. I really just intended to throw out some questions that interested me while reading her excellent piece.

Reading back over my post, I confess that I confused two separate issues. One is the question of whether Kelly-Hawkins' race should remain an open question. The other is the question of what evidence is sufficient, not only in her case but in any case, to declare a question about race "closed."

In Kelly-Hawkins' case, I agree that the preponderance of evidence suggests she was always identified as white -- by herself, by her family, apparently by her contemporaries, and even (setting aside for the moment Tim Burke's objections to a blunt historicism) by her novels. All of this, taken together, does suggest what everyone else has already been suggesting: that the strange career of Kelly-Hawkins has had more to do with the scholarly project of constructing an African American literary canon than with the historical evidence itself.

Since I agree with you on all these issues, what I really wanted to raise in the post was Issue #2: how particular kinds of evidence (in this case, the census) are used to make fluid and historically contingent categories like race more fixed than they actually were. I don't mean to endorse a severe skepticism about the census, only a moderate one. And I was concerned (perhaps wrongly) that Jackson referred to the census uncritically as the final silencer of skepticism about Kelly-Hawkins' race. I think we can grant she's right (given all the evidence) about this case, yet still urge that the census be approached with more caution, given its own strange career on matters of race.

I also apologize if I came across as snarky in my reference to your post. I definitely didn't mean to treat lightly your very serious questions about Gates. 

Posted by Caleb

Posted by Anonymous Anonymous on 3/02/2005 06:22:00 PM : Permalink  

P.S. I've pointed out, in brackets, some of the sloppy sentences in my original post -- points where I was confusing the two issues I mention in the above comment. 

Posted by Caleb

Posted by Anonymous Anonymous on 3/02/2005 07:00:00 PM : Permalink  

Caleb, I think we're in agreement and as far as I am concerned you are the soul of graciousness. I've seen snarky and, unless it was me, it isn't here. My current project in Virginia, btw, involves a bi-racial family whose members have different racial identities from census record to census record, so I agree fully that they need to be supplemented with other kinds of evidence in order to get a more accurate picture of what the situation actually was. 

Posted by Ralph Luker

Posted by Anonymous Anonymous on 3/02/2005 07:21:00 PM : Permalink  

I found Emma Dunham Kelley in 2003 and have been researching her ever since. I was literally proofreading my footnotes to a paper I am submitting to the peer-reviewed "National Genealogical Society Quarterly" on Emma's life and entry into the AA canon.

Katherine Flynn 

Posted by Katherine Flynn

Posted by Anonymous Anonymous on 3/03/2005 07:31:00 PM : Permalink  


I was literally proofreading my footnotes on my paper on Emma when Ms. jackson's article appeared. I hope to have my research out soon so that the whole case can be seen and evaluated.

Katherine Flynn

Posted by Katherine Flynn

Posted by Anonymous Anonymous on 3/03/2005 10:23:00 PM : Permalink  

This makes me desperate to know what Katherine Flynn's article is going to say when it comes out.

But whatever it says, I completely agree with Caleb that in many ways to respond to this issue by trying to determine factually whether she was or was not a particular racial group is to kind of miss the point. The rich literature on the existence and consequence of "passing" in American history is only one place to look to gain an understanding of the futility of trying to make firm and final judgments now about the exact racial identity of people in the past.

This is one reason why, whatever my criticisms of Ward Churchill, I wouldn't for the life of my dream of trying to debate whether he is or is not a real Native American--the fiendishly complex layering of claims, standards, and laws used to make that determination over time mean that such a claim is always social and rhetorical, never cooly empirical. You can certainly note the irony of Churchill's own writings on the subject of racial and cultural authenticity in relation to his own ability to inhabit that authenticity within the terms that he describes, but even that irony is hardly limited to him or to the present. The most vehement practicioners of identity politics tend to be those whose claims to a particular identity may be considered contentious or marginal.  

Posted by Timothy Burke

Posted by Anonymous Anonymous on 3/04/2005 02:37:00 PM : Permalink  

Thanks for that comment. I couldn't have said it better, but, throwing caution to the wind, I've kept talking about this in a post at Cliopatria , which also includes some links to other posts on this subject that readers might find interesting. 

Posted by Caleb

Posted by Anonymous Anonymous on 3/05/2005 12:04:00 PM : Permalink  

I have been enjoying all of this chit chat about Emma Dunham Kelly-Hawkins. My grandmother, Charlotte Meader Moulton was the daughter of Alice Kelly Meader, Emma Dunham-Kelly's sister. When Emma's daughter, Megda, passed away, she left the contents of her home in Rumford, R.I. to my uncle. I spent 3 months in her home cleaning it out and I do have several of Megda's diaries from the 1920's and 1930's. Emma Dunham Kelly-Hawkins was my great-great aunt. I have been curious as to the information "Holly" obtained as to a relative she spoke with in regards to Emma. I believe she spoke after she wrote the article with a descendent of Isaac Kelly, but she was not directly related to Emma. I would be happy to speak to anyone regarding my great-great aunt, Emma Dunham Kelly-Hawkins. My uncle has done great research on his mother's family and it is quite interesting.  

Posted by Carol Moulton Isacco

Posted by Anonymous Anonymous on 3/21/2005 01:25:00 AM : Permalink  

Please note: RE: Emma Dunham Kelly-Hawkins.
I can be reached at e-mail:
I am willing to speak with anyone regarding this relative of mine.  

Posted by Carol Moulton Isacco

Posted by Anonymous Anonymous on 3/21/2005 01:29:00 AM : Permalink  

I have received calls regarding Emma Dunham Kelly-Hawkins (my great-great aunt). If your main concern is whether or not she was African-American, she was not. I can go back to Emma's parents and her grandparents, none of which were African-American. Would it not be plain common sense to say that every Caucasian may have an ancestor who may have been African. How far do you plan to go back in time to find out this information? I do wish that Emma could be here to enjoy her fame. 

Posted by Carol Moulton Isacco

Posted by Anonymous Anonymous on 8/07/2005 06:47:00 PM : Permalink  

Wow. Carol Moulton Isacco seems to have closed the book. Now what to do with my once apparaised $5,000. 4 book collection of 1st and 2nd editions!

What was interesting before, has become even more interesting. I assume that is why Barnes and Noble is no longer listing the pricey values for her dated, unstated first ed. and second edition.

Posted by Anonymous Anonymous on 12/11/2005 06:18:00 PM : Permalink  

Post a Comment

Back to Main Page

Site Meter