Essays

doi:10.17613/4y30-th10 | PDF

In creating a project to investigate the relationships among members of the Belfast Group, it is important to know exactly what that Group is. Being specific about this when creating our data was critical so we could accurately measure who was connected to this thing we call “the Belfast Group.” But, as often happens with humanities data, it turns out that things are a little messy. In this case, while the term originally refers to the writing workshop begun by Philip Hobsbaum, many critics and commentators have also used it to refer to the idea of a Belfast “school” of poets (see Clark 1, 6). Many members of this supposed school—Seamus Heaney and Michael Longley, among others—were, of course, participants in the writing workshop, which adds to the slippage between the two uses. But while it is demonstrably true that a writing workshop existed, it is less clear whether there was any unified purpose that might constitute a school; as Norman Dugdale put it, the “The Group had no manifesto, no corporate identity, no programme beyond providing a forum in which writers […] could produce their wares and have them discussed” (Dugdale et al. 54). For the purposes of this site, then, when we speak of the “Belfast Group,” we mean the weekly writing workshop founded by Hobsbaum and continued by Seamus and Marie Heaney, along with Michael Allen and Arthur Terry.

Even having made it clear what we mean when we say “Belfast Group,” it’s also true that the writing workshop was not a single, consistent entity. Split into two phases and involving a rotating cast of writers—some of whom deny their involvement or its significance—the Group was anything but the fixed entity that would be desirable for conventional data analysis. And the multiplicitous nature of the workshop was more than just membership. Individual accounts of how the Group worked or who was involved regularly contradict each other. Trying to account for all of these varying perspectives fifty years later, it is even more difficult to know what the Group was and how it functioned. Indeed, given these conflicting memories, one of the goals of this project is to create an alternate way of considering what the Belfast Group was. This data-centric representation of the Group is, of course, not “Truth,” with a capital T, but simply one more perspective, as we write in our essay about archival bias.

What follows here is a brief account of the Group members’ singular perceptions of the writing workshop and some of our thoughts about doing data work on an entity which is anything but singular.

Group Think? Conflicted Reminiscences of the Workshop

When trying to understand the Belfast Group, it seems logical to go to the source: the people who participated in it. Yet, in a collection of memories about the group—termed a “Symposium”—published in the Honest Ulsterman in 1976, we find a number of conflicting descriptions of the workshop. Longley described being “rather surprised by the ferocity of Hobsbaum’s attack” on his work, but added that he eventually “look[ed] forward masochistically to the seasonal maulings” (Dugdale et al. 57). Jack Pakenham recounts enjoying the “verbal battles which were part and parcel of the meetings” (Dugdale et al. 58), and Derek Mahon claimed that the Group met to “read and savage one another’s work” (“Poetry in Northern Ireland” 90). Simmons, on the other hand, recalled that “all there was kindness and encouragement,” a fact which “irritated” him as it diluted any potential feedback (Dugdale et al. 59). Bernard MacLaverty, in a 2000 interview, suggested that Hobsbaum would always “find something to say to validate what was written on the page” (qtd. in Clark 44). In another venue, Joan Watton Newmann wrote that “the ethos [Philip Hobsbaum] created and sustained was one of pleasure and discovery. There was no space for destruction” (“Coming of Age” 118). Arthur Terry split the difference between the camps, calling the workshop’s tone “an agreeable mixture of friendliness and astringency,” as did Seamus Heaney, who wrote that if Hobsbaum “drove some people mad with his absolutes and hurt others with his overbearing, he confirmed as many with his enthusiasms” (Dugdale et al. 61, 62).

In addition to disagreeing about the general tenor of meetings, the Group participants also express differing opinions about the types of poetry that were favored at the meetings. Both Pakenham and Longley comment on a particular aesthetic bias that ruled the group. As Longley put it, “Hobsbaum’s aesthetic demanded gritty particularity, an unrhetorical utterance” (Dugdale et al. 56). This mode conflicted, he notes, with both Pakenham’s “free-wheeling surrealist verse” and his own disposition “as a lapsed Classicist” (Dugdale et al. 56). Terry, on the other hand, writes that “nothing approaching a ‘group-mentality’ ever emerged” given “the sheer range of opinions and the presence of so much individual talent” (Dugdale et al. 61).^[1] Perhaps, then, Pakenham and Longley noticed the dominant aesthetic because they found themselves so far outside the prevailing style.

The members also had differing perspectives on the mix of people involved in the workshop. Pakenham thought the group was too academic and “University orientated [sic],” because “the majority of the Group [...] were either University Lecturers or their students” (Dugdale et al. 57). But Arthur Terry felt that one of the Group’s strengths was its “being a meeting-place for people of very different backgrounds and interests, many of whom were refreshingly unconnected with the University” (Dugdale et al. 61). Since Terry taught at the university, the inclusion of anyone from outside that world could have felt like a breath of fresh air; for Pakenham, who worked at Ashfield Boys’ High School as “an English teacher involved in the teaching of Literature,” the number of university-affiliated members must have felt conspicuous (Dugdale et al. 58; see also Gormley’s Art Auctions and Culture Northern Ireland).^[2] Yet this difference in opinion might not come down to different perspectives: Terry and Pakenham might also be remembering the Group at different points over the course of its nine years. The Group really might have been different given its separate phases and conveners and what Hobsbaum called the “turnover of talent” (Dugdale et al. 55). For example, Michael Longley writes that he “never saw Simmons at a Group meeting” (Dugdale et al. 56). Thus, while Heaney commented in 1976 that the Group allowed the public to think of the writers as a “single, even singular phenomenon,” this seems to be a kind of useful fiction (for both public and writers), and was clearly an oversimplification even at that point in time (Dugdale et al. 62).

Whether or not the writing workshop was dominated by people affiliated with Queen’s University, most individuals haven’t questioned their participation in the Group. But that is precisely the case with Derek Mahon. In 1970, Mahon wrote an article in Twentieth Century Studies about the development of “Poetry in Northern Ireland.” He mentions the “group seminar which met weekly in [Hobsbaum’s] flat” and suggests that “The Hobsbaum seminar (known as ‘the Group’...) was probably the first to crystalise the sense of a new Northern poetry” (“Poetry in Northern Ireland” 90, 91). In continuing, Mahon clarifies the importance of the Group for the authors in Northern Ireland: “Here was this man from London, people thought, whose name and whose friends’ names appeared in leading journals, and he’s actually taking us seriously” (91, original emphasis). Perhaps it is the “us” in “he’s actually taking us seriously” or the fact that Mahon writes so knowingly—if briefly—about the Group here, but many people associated him with the workshop. This could have also come through his close friendship with Longley.^[3] Longley was clearly a member of the writing workshop—although not a founding one, despite Hobsbaum’s occasionally listing him as such^[4]—and it therefore stood to reason that his compatriot Mahon must be as well. Yet Mahon claims to have only gone to a single meeting of the Group and has since vehemently denied playing any real part in it, as he did in a 1991 interview: “I was not a member of Philip Hobsbaum’s fucking Belfast Group. I was in a different city. I was a member of my own group in Dublin. I went once to Philip’s group, and never again” (“Q. and A.” 28, original emphasis).^[5] While he might have attended only a single time, we do have one Group sheet of Mahon’s poems in the collections at MARBL, which suggests that his poems were discussed on one evening. At first glance this seems to suggest that Mahon did more than just drop by the meeting, as he likely would have had to schedule his attendance and make the poems available to the Heaneys, Terry, or Allen ahead of time for distribution. But as Stephen Enniss discusses in his recent biography of Mahon, “a number of the poems present on the sheet were not written until after Mahon had left Belfast for Canada” (269 n. 62). Enniss suggests that Longley probably “took it upon himself to prepare a Group sheet of Mahon’s poems that he presented in his place” (270 n. 62).

Yet Mahon may have been more inclined to be considered part of a group that was making some literary noise earlier on, at the time of his article in Twentieth Century Studies. He describes Northern Ireland therein as a “cultural desert” in the 1950s and, in the opinion of many, “traditionally philistine” (“Poetry in Northern Ireland” 90, 89). Heaney suggests that Hobsbaum’s ability to energize the Belfast literary scene was in part because he had “trust in the parochial, the inept, the unprinted” (Dugdale et al. 62). If these two descriptions are accurate, one could imagine that inclusion in a group that had some literary recognition would have been valuable to young poets. But, as Heather Clark argues in The Ulster Renaissance, once the individual authors’ reputations were on the ascent, it would perhaps be less desirable to be seen as members of a “school” (173–207 passim.). Indeed, while Mahon is the only person who challenges his association of the Group, many others have downplayed its importance for—and, by extension, Hobsbaum’s influence—on their writing. Longley, for example, writes that “I can honestly say that I didn’t alter one semi-colon as the result of Group discussion” (Dugdale et al. 57).^[6] Pakenham believed that the excellent poems and stories he heard at the group were written “in spite of the criticism expressed there not because of it” (Dugdale et al. 58). Others declined to comment in The Honest Ulsterman’s “Symposium” “on the grounds that they regarded the meetings as little more than pleasant social occasions”—at least as they looked back from the more established literary moment of 1976 (Dugdale et al. 53).^[7] Yet Mahon, of all people, wrote that Hobsbaum’s “enthusiasm generated activity in people who might otherwise have fallen silent” (“Poetry in Northern Ireland” 91).

Regardless of what the Belfast Group was, how it worked, or who felt that it was important, everyone appears to agree on one thing: the reason for being there. As Stewart Parker writes, “What did we find to talk about so interminably[?] [...] [W]hy, the world was young, poetry was under every stone, just waiting to crawl out and be mimeographed. Poetry! God how we craved it, molested it, exalted it, lived it—but above all explicated it!” (Dugdale et al. 59).

Memories vs. Data

It’s important to remind ourselves of the difficulty of knowing what the Belfast Group was when analyzing it from a data-centric point of view. After all, visualizing the relationships among the Group and its members requires us to treat it as a single entity with a stable identity. We seem to confirm this thinking, whether it is meaningful or accurate, when we create an entry for the Group itself in a dataset or use a linked-data URI for reference. The neatness of a single-line entry for the writing workshop seems to contradict the contradictory recollections of its many different participants.

A desire to overcome this all-too-singular thinking is part of the motivation for creating the visualizations on Belfast Group Poetry|Networks. The graphs and maps do not seek to resolve the incompatible reminiscences of the Group so much as present one more lens through which to view its activities and thereby disrupt the notion of a static thing that was the “Belfast Group.” For example, one of our network graphs attempts to fracture the concept of a stable Group by highlighting the Group during its two separate time periods. This visualization makes it easy to see that there really were two different groups of people participating in the workshop. Individuals like Stewart Parker, Joan Watton Newmann, Brian Scott, and J. K. Johnston only participated in the first instantiation, and Ciaran Carson, Paul Muldoon, and Trevor McMahon, among others, attended only when the Heaneys were leading the Group. This graph also helps us identify those individuals who span both phases, among them Seamus Heaney, Michael Longley, and James Simmons, but also Iris Bull, Maurice Gallagher, Norman Dugdale, and more. A version of this network graph is presented below in a static form so as to make it easier to read. We have also colored the different nodes on the graph according to algorithmic community detection, where the computer identifies subsets of nodes that are tightly connected to each other.

Of course, it should be said again that this visualization is incomplete. Since it only represents those who authored or owned Group sheets from this particular period, many individuals do not appear. Marie Heaney, for example, was clearly a participant in both phases of the workshop, but she does not appear in this visualization due to how the Group worked and due to cataloging practices (both of which we address below). Similarly, Brian Scott might have attended during the second phase but simply not authored or owned any Group sheets that have made their way into the collections at Emory or Queen’s.

By transforming our data in other ways, we’re able to gain insights into the Group members’ relationships with one another. As mentioned above, the influence of the workshop on the participants’ writing is questionable; Michael Longley, for example, provocatively wrote that “the poetry would have happened anyway” (Dugdale et al. 57). While it is impossible to accurately tease out questions of influence or might-have-beens, we can use it to look at the various descriptions of the people involved with the Group as a way of seeing how well our data matches up against the memories of the people involved.

For instance, Norman Dugdale discusses some of the most important members of the workshop in the “Symposium”: “In my time, Hobsbaum, Heaney, Longley and Terry formed the core of the group” (Dugdale et al. 54). He goes on to say—using a poetic metaphor not inappropriate for network theory—that these four were “circled by many lesser moons and a shifting penumbra of casual attenders” (Dugdale et al. 54). How does this recollection comport with our data? Looking at our visualization of Belfast Group authors by period, which is based on the authorship and ownership of Group sheets with inferred dates, we find that Hobsbaum, Longley, and Heaney are in the top four nodes by almost every network measure that might indicate the importance of influence of a node in a network. These measures include the total number of connections that each of them has to other people (“degree”); their likelihood of being along the shortest path between any two points on the network (“betweenness centrality”); and how many connections the nodes that they are connected to have (“eigenvector centrality”). This last measure is often used as a proxy for understanding influence in a network and resembles the PageRank method that determines Google’s search results. In each of these measures, Hobsbaum is by far the highest, which can be attributed to his being the sole owner of the sixty-six Group sheets at Queen’s, which make up just under fifty percent of our data set (n=138).^[8] In this sense, our data for Hobsbaum, Heaney, and Longley matches Dugdale’s recollection of the Group...except for Arthur Terry.

Throughout our data, Terry—then a professor of Spanish at Queen’s and noted translator of Catalan works—is relatively unconnected to other members of the Belfast Group. First, since we do not have his papers in MARBL, we do not have record of either his correspondence or his ownership of Group sheets.^[9] Having access to this information would have caused him to appear more prominently in our visualizations of people associated with the Belfast Group and the Group’s authors by period, respectively. What’s more, we only have records of three Group sheets that he authored, all of them in the first period of the Group. Without a paper trail, then, he becomes relatively invisible in our measures, despite Dugdale’s pointing to his role in the “core of the group.” More vexing, in our overview of the Belfast Group, Enniss notes that Terry was a co-organizer of the second phase of the Group, along with the Heaneys and Michael Allen (“The Belfast Group Writing Workshop”).^[10] (Michael Allen, it turns out, is even less represented in our data than Terry.) A way to reconcile our data and Dugdale’s recollection, then, is the likelihood that Terry became central to the Group by his bringing people together and attending meetings but did not author any Group sheets—or at least any that we know of—once Hobsbaum left.^[11]

If we turn to the relationships clustered around James Simmons, we have another opportunity to compare our data with individuals’ perceptions about the Group. As noted above, Simmons reported after the fact that he found the feedback at the workshops to be too positive, on the whole. Still, he had “no doubt that those evenings were useful educational events” and “was glad to have some serious people paying attention to [his] work” (Dugdale et al. 59). In particular, he calls attention to Arthur Terry for being “particularly lucid and balanced as a critic”—again pointing out Terry’s importance to the Group, despite what our data may seem to indicate (Dugdale et al. 59). Yet Simmons writes that his “memories of the whole thing are very vague” and, perhaps most germane for our purposes, claims, “I never really got to know any of them well, perhaps because I was that much older” (Dugdale et al. 60). Indeed, when we look at Simmons’s connections in our visualization of people associated with the Belfast Group, which represents the relationships among individuals and organizations connected to the workshop, we see that he is a bit more isolated than other equally well known members of the Group. While Simmons has connections of varying strength with Michael Longley, Derek Mahon, and Paul Muldoon thanks to his correspondence with them, he lacks any connection whatsoever with Hobsbaum, Edna Longley, or Ciaran Carson. This of course does not mean that they didn’t know Simmons, but that our data does not account for the nature of their relationship. This lack of direct connections to some of the key members of the Group results in Simmons being excluded in our data model from a number of different cliques within the Belfast Group, where a “clique” is defined as a subgroup in a network where every individual is connected to every other individual. Whether or not it was Simmons’s age, that he lived in Nigeria from 1964–67, or that when he did live in Northern Ireland it was almost always outside of Belfast, our model suggests that Simmons does seem to have “never really got to know” the members of the workshop as well as some others did, although he did edit some of their work during his time at The Honest Ulsterman.^[12]

We can see this disconnect play out more specifically in the relationship—or lack of one—between Simmons and Michael Longley. As already mentioned, Michael Longley states in the “Symposium,” “I never saw Simmons at a Group meeting” (Dugdale et al. 56). Again, our data appear to corroborate this statement. When considering our visualization of Group authors by period, which represents authors and owners of Group sheets, we see a relatively faint connection between the two authors. This comes from there being just a single Longley Group sheet in the Simmons papers and none whatsoever authored by Simmons in the Longley papers. This is somewhat surprising as Longley’s ten Group sheets make up the second largest total for a single author in the data set, and Simmons is tied for fourth with four Group sheets; what’s more, the Longley and Simmons papers contain twenty-seven and eighteen Group sheets, respectively. Since the two men don’t own more of one another’s work in this format, there’s a strong likelihood that they did not attend the same Group meetings. Also plausible, of course, is the fact that one or both of them simply didn’t keep the other’s work. If Longley did not see Simmons at Group meetings, there is a good chance that others did not connect with him either. That said, the two men are connected in the visualization of people associated with the Belfast Group by virtue of their correspondence, even though much of it takes place after the conclusion of the Group.

It is similarly enlightening to consider Simmons’s relationship with Hobsbaum. While no correspondence exists in MARBL between Simmons and Hobsbaum, the Simmons papers do contain one Hobsbaum Group sheet and Hobsbaum’s papers at Queen’s have three Simmons Group sheets. The result is that while there is no connection between the two on our graph of people associated with the writing workshop, a clear relationship between them appears on our graph of authors and owners of Group sheets by period. This marks in some ways the reverse of the Simmons–Longley relationship, where considerable correspondence exists but almost no Group sheets. The two graphs measure different things and point simultaneously to the limitations of a data-centric approach—the fact that Simmons clearly had a relationship with both Hobsbaum and Longley that is not completely captured by our data—and this approach's strengths—the fact that these relationships seem to be of a different quality.

While Simmons claims in the The Honest Ulsterman—a publication that he founded—that he didn’t really remember the Group meetings nor know the participants well, he also undermines that claim in the same paragraph: “I seem to remember preferring some of Joan Newman’s [sic] poems (Joan Watton she was then and lodged with us in Lisburn). [...] Stewart Parker and Hugh Bredin I remember too” (Dugdale et al. 60).^[13] Our visualization of Group sheet authors and owners indeed shows a connection between Simmons and these three poets. In each of these three cases, the edge between Simmons and the other poet is the same weight because there are two Group sheets from each of them in Simmons’s papers in MARBL. Due to this, one might reasonably assume that Simmons had the same sort of relationship with Parker, Bredin, and Watton Newmann. Yet Simmons’s comments in the “Symposium” suggest that things should perhaps be visualized differently: while he “remembers” Parker and Bredin, Watton Newmann lodged with Simmons and his wife Laura Stinson. That surely suggests a stronger relationship, as do his further comments: “Joan Newman [sic] I knew and loved, the girl and her work, and so that is what remains with me of that period. The critical evenings I never think of; but the excitement of reading her poems as they came out is with me still” (Dugdale et al. 60). From the vantage point of 1976 when The Honest Ulsterman took his comments on the Belfast Group, it seems clear that Simmons felt he had no connection as strong as that with Watton Newmann. Yet, based on how our network data was collected, there was no way we could know about or include this external, affective connection between the two. So while our approach confirms some of what Simmons reports, it belies other—perhaps more important—parts of his recollections.

Conclusion

The examples in this essay hopefully point to how a data-centric approach can clarify our understanding of literary communities. Such a method allows us to verify claims made by members of the Group about who was there and what roles they played. It helps us tease out the different time periods within a protean group such as Hobsbaum’s writing workshop, and it helps us better conceive of the range of possible entities that could fall under the rubric of “the Belfast Group.” At the same time, the recollections of Group members point to places where our data are incomplete and could, in all likelihood, never be complete. That said, we can imagine other questions to ask of the data: for example, examining the strength of relationships among what Hobsbaum calls the “founder members” and the relationships these founders form with “later entrants” (Dugdale et al. 55), or, taking a page from Simmons, looking at the connections among poets of different generations, such as those like Ciaran Carson and Paul Muldoon who joined the Group in its second iteration. It’s possible that no meaningful connections will be found when asking these questions, and that’s an outcome that should be familiar and acceptable to both literary scholars and digital humanists. But patterns will occasionally surface that scholars can interpret, and that chance makes it worth the pursuit.

This essay was peer reviewed by Geraldine Higgins and Nathan Suhr-Sytsma.

1. ^ In an essay about the epistolary poems of Michael Longley and Derek Mahon, Gavin Drummond writes, “While the Group may or may not have affected the aesthetic of these writers, their informal groupings did. [...] The poets in this community may have disagreed with each other poetically and politically but they listened to each other, habitually communicated with each other, and often changed their poetry on the advice of one another” (32).
2. ^ While Belfast was a relatively small community, Queen’s University made up an even smaller portion of that population. The population of Belfast in 1961 was 415,856; by 1966 it had actually declined to 399,270 (Government of Northern Ireland 6). Contrast this with the population of London in 1961: 3,200,484 (1961 Census of England and Wales). In 1966, there were a total of 5,371 students (full-time, part-time, and postgraduate) enrolled at Queen’s, and there were 562 full-time academic staff—88 of them in Arts—and 78 full-time lecturers (Clarkson 195-212 passim.). Any literary community in Belfast and surrounding Queen’s would have been smaller still. Perhaps the grouping of disparate talents into a “school" can be attributed to this environmental scale, in which authors were likely to be in contact with one another.
3. ^ In The Ulster Renaissance, Heather Clark points to the friendship of Mahon with Longley, as well as with Heaney, as “forever link[ing] him with Hobsbaum’s Group”: “Heaney, Mahon, and Longley formed a subset of that Group, which Michael Foley referred to as the ‘Tight Assed Trio.’ These three gave more readings together in Belfast during the 1960s than any other Group configuration, cementing the public’s perception of them as a clique” (57).
4. ^ In a 2000 interview with Heather Clark, Hobsbaum lists Michael Longley among the “later entrants” to the Group while identifying Edna Longley as an original member (Clark 55, 54). Yet in The Honest Ulsterman’s “The Belfast Group: A Symposium,” which was published in 1976 and therefore much closer to the events at hand, Hobsbaum names “Michael and Edna Longley” among the “founder members” (Dugdale et al. 55). Hobsbaum’s later account along with Michael Longley’s own recollection—“The group was a going concern by [the time I attended]: I can in no way be seen as a founding member”—suggests that the former was simply mistaken in The Honest Ulsterman (Dugdale et al. 56).
5. ^ Mahon’s comments about his presence at the Group are reported as fact by Frank Ormsby in the introductory remarks to “Symposium” in The Honest Ulsterman (53). Stephen Enniss’s biography of Mahon repeats the claim, based on his schedule of courses at Trinity and his moves to Paris and Canada in 1964 and 1965, respectively (57, 61–62). Finally Hobsbaum, in an email to Clark in 2005, stated that Mahon attended the Group just once “when he was back in Belfast” (55). The claim is a little less precise when reported by Longley in the “Symposium,” however; he writes that “Mahon was present only once or twice as a kind of outside observer when he happened to be back in Belfast and staying with me” (56, emphasis added). While the possibility of Mahon’s attending more than once is intriguing, in the end, it is not critical to know exactly how many times that Mahon attended the Group. Instead it’s simply worth noting (again) that humanities data are tricky, as they so often consist of memories and paper trails.
6. ^ In a “Recollection” about the Belfast Group, Hobsbaum writes, “The tendency to use the Belfast Group as a means of revision was not general” (178) which corroborates Longley’s statement.
7. ^ While some individuals declined to comment for The Honest Ulsterman, it is very likely that not everyone who participated in the Group was invited to do so. In his editorial comments Frank Ormsby does not make it clear whom he contacted, but it is safe to assume that some of the less well known or less published participants—Iris Bull, Hugh Bredin, Lynette M. Croskery, or Joan Watton Newmann (about whom more below), among others—probably were not asked. It is especially interesting to note that no women contributed to the “Symposium.”
8. ^ When calculating these same network measurements against the complete data set (as represented in our visualization that shows individuals connected to the Belfast Group, including second-degree connections), both Longley and Heaney remain ranked in the top eight individuals in measure of degree, betweenness centrality, and eigenvector centrality. Hobsbaum, on the other hand, falls considerably in every measure. This can be attributed to the fact that the complete data set relies heavily on the collections in MARBL which holds the bulk of Heaney’s and Longley’s papers, but only a small box of Hobsbaum’s materials. Our approach could be expanded to include materials at other archives—even if they did not natively expose their data in appropriate formats. Even if we had the time and resources to do this work, Hobsbaum’s papers have not yet (as of this writing) been processed at the University of Glasgow.
9. ^ The fact that Terry’s papers have been collected neither in MARBL nor anywhere else, including the University of Essex where he worked for 21 years (Philip Terry), says something again about the biases inherent in archives. Since Terry was a scholar and a translator, his output is seen as less valuable and interesting to most archives.
10. ^ Terry’s obituary in The Guardian corroborates both Dugdale’s and Enniss’s comments about him: “In the Belfast Writers’ Group, founded in 1963, his original contributions were limited to verse translations, but his presence was indispensable, as was his organising work. A valued friend to poets since his Barcelona days, he now won esteem from Philip Hobsbaum, Seamus Heaney, Michael Longley and many others” (Round).
11. ^ A recent email from Arthur Terry’s son Philip to the authors helps establish that Terry was indeed in attendance in the second half of the Group (Philip Terry). Philip Terry included scans of three Group sheets (two of them identical) enclosed in his father’s copy of Seamus Heaney’s Death of a Naturalist (1966): two copies of the Heaney sheet that starts with “Icon” and one copy of the sheet with “Servant Boy.” One of these Group sheets carries the same handwritten note (subsequently duplicated) and about the date of the workshop as appears on copies of the Group sheet in the Simmons and Carson papers in MARBL. Since Carson did not participate until the second phase of the Group, we have clear evidence that Terry did as well. If these Group sheets were added to an archival collection and made part of our data set, Terry would consequently be connected to the second phase of the Group.
12. ^ Clark writes extensively about how Simmons perpetually felt excluded from the Belfast poetry scene, both during and after the Belfast Group workshops; see 87-89, 176-82 passim.
13. ^ In The Honest Ulsterman, Simmons refers to Watton by her married name, Joan Newmann, although he spells it incorrectly as “Newman” (Dugdale et al. 60). For consistency’s sake, we use her full name to refer to her in our essays, since her maiden name appears on her Group sheets.

doi:10.17613/jzcc-nt41 | PDF

Archival Biases

Social network analysis is typically used where data are complete and all connections within a system are known. However, as other humanities networking projects have discovered, building a network based on historical data means that we are inevitably working with incomplete information. In other words, the lack of connections in our graph doesn’t mean that no connection exists, but only that we have no documented evidence of one. For a large-scale, historical project like the Republic of Letters, this incomplete information is due to the historic nature of the content they are working with; in other words, not all of the evidence exists any more (see Chang).^[1] In our case, we have a different bias and different missing data because we have primarily used information from a single archive as the source of our network data.^[2]

Emory’s Manuscript, Archives, Rare Book Library (MARBL) has a strong collection of Irish literary materials, which is part of what inspired this project in the first place, but certainly there are other archives in Ireland, the U.K., and the U.S. with other materials that would be relevant to this project. For example, Philip Hobsbaum’s papers are held at the University of Glasgow, where they are still being processed. Including information about his correspondence would without a doubt increase the number and strength of his connections, as well as introduce other individuals into the orbit of the Belfast Group. As linked open data becomes available in the future, it should be possible to incorporate data from additional archival collections into a project like this.

Even beyond the fact that our data is limited by the selection of collections we’ve pulled correspondence information from, our data and network models are biased simply because they are based on data from archival collections. As discussed in our essay on women in the Belfast Group, certain types of people tend to be more prominent in archival collections than others—namely, famous authors. Our network is based on what archives have collected; what archivists have determined is worth keeping; and how much detail those archivists have used to describe those materials. Some manuscripts collections in MARBL are described in much greater detail than others. For example, prior to this project, all the Group sheets in the James Simmons papers were described collectively as “Belfast Group sheets.” To generate more accurate data, we created item-level descriptions of the individual sheets. Similarly, the level of detail in the description of an author’s correspondence varies widely. Our data are incomplete for collections—such as the Derek Mahon papers—where the processing archivists chose to mention correspondents that they considered notable rather than explicitly naming every individual. Observing the gaps in our data is not to impugn the work of our archivists; processing a collection is extremely labor intensive and a complete description of materials would ultimately border on unwieldy facsimile, as the fable from Borges makes clear.^[3] What’s more, the conditions of a gift to an archive—the terms of which are generally not disclosed—can influence the extent of the descriptions.

The biases inherent in our archival-based data are more evident when looking at the full network generated from our data, especially when algorithmically detected communities are highlighted.

We generally see distinct communities for the major figures corresponding to specific MARBL collections, such as Paul Muldoon, Peter Fallon, Tom Paulin, and Ted Hughes; even emeritus Emory faculty member W. Ronald Schuchard has a small community distinct from the rest of the network. The number of distinct communities is most likely the result of the large number of correspondents detailed in each of these collections who do not appear in the other collections. Heaney, Longley, Mahon, and Carson are much more centrally connected in this network representation. While literary historians like Heather Clark would certainly place Heaney and Longley at the center of narratives about the Group, they appear centrally here for another reason: the efforts of our library to collect materials from those connected to the Belfast Group. Our data, again, reflect our collections as much as they reflect reality. When considering the community that the algorithm places the Belfast Group within, one finds several members of the workshop: Longley, Carson, Simmons, Ormsby, Hobsbaum. This detail surprised us given the fact that, as we noted on in our previous essay, there are no connections between Simmons and Carson on the network graph of those associated with the Group,.^[4] The lack of connection is perhaps explained by Simmons’s more frequent participation in the initial gatherings under Hobsbaum, while Carson, who is Simmons’s junior by 15 years, could only participate in the second half of the Group.^[5] Yet, the algorithm placed them in the same community, in large part thanks to their connection to some of the same correspondents. Although this placement is algorithmic, it might also be explained by them both spending almost their entire careers in Northern Ireland, as opposed to Heaney or Muldoon, who worked in the US and other countries. By virtue of remaining in the same environs, Simmons and Carson have a smaller circle of correspondents—at least in our data—which makes it less likely for them to be sorted into another community.

Archival Futures

The limitations and biases that we’ve described above are simply the nature of archives, and researchers need to be aware of them when doing any sort of archival work, including data-driven approaches. With these caveats in mind, a network approach to archival data can still reveal new things about literary histories. What’s more, a network approach could transform how we use the archive itself.

By applying the same methodology taken for this project, we could enhance data in finding aids from all of MARBL or another large archive. We could then use the resulting data to get a broader sense of the networks within the full archive. MARBL’s collection strengths in Irish literature, African American literature and culture, and southern history, would likely appear on a larger scale when we look for communities within the full network of people represented by the archive. But we might also find unexpected connections bridging those communities that would otherwise escape the notice of most researchers or even individual archivists. Visualizing the network of the entire library could be a useful tool for scholars and researchers, enabling them to find related materials in other collections and get a better sense of their research subjects within the context of the archive. Indeed, such an interactive visualization might go a long way to replacing the finding aids as the primary interface to an archive’s materials. If multiple institutions began using linked open data in this manner, it would even help a researcher identify other collections in which she or he might look for related materials. Creating such a network could also be a helpful tool for archives to give potential donors a beautiful, big-picture view of the collections in the archive and how their materials might connect.

This essay was peer reviewed by Geraldine Higgins and Nathan Suhr-Sytsma.

1. ^ Of course it is worth noting that there are gaps in the historical record of events of the Belfast Group, which took place only 50 years ago. For example, while our data set contains information about all the Group sheets that are known to exist, we almost certainly do not know the location of all them. Any of the participants could have kept the sheets from the meetings they attended. We have recently learned, for example, that Arthur Terry kept a couple of Seamus Heaney Group sheets that have been uncollected in any archive (Terry).
2. ^ There are many reasons why our information primarily comes from one archive. The first and most obvious is that since we work at Emory, we have easy access to the documents here. But more importantly, we had the ability to make changes to how the Emory finding aids expose data. We draw directly from the enhanced finding aids to generate the data for the site. By contrast, the finding aids for the Belfast Group materials at Queen’s University Belfast are a PDF. While we have incorporated those data, since this project is about enhancing and re-combining library data, working in our own environment is most conducive to demonstrating what is possible. We hope that other libraries will begin exposing their collections in similar ways.
3. ^ Jorge Luis Borges. “Museum: On Exactitude in Science.” In Collected Fictions. Trans. Andrew Hurley. New York: Viking, 1999. 325.
4. ^ There is a direct connection between Carson and Simmons in our network graph that visualizes Belfast Group authors by period and shows relationships that we have inferred among Group members based on Group sheet ownership. Carson and Simmons both owned the same Heaney Group sheet. This suggests the following possibilities: 1) they both attended these meetings but didn’t know one another or form a relationship that would lead to correspondence or other connections; 2) one or both of them received a mailed copy of the Group sheet but did not attend the meeting in question; or 3) one of them got the Group sheet through some other means. Since the Simmons papers have seven Group sheets from the second period of the Group, it seems likely that he did attend but did not, for whatever reason, end up forming a relationship with Carson that led to their correspondence of such significance that it would be called out in the MARBL Finding Aids. This lack of correspondence, along with their not being described as colleagues or knowing one another in the finding aids, results in Carson and Simmons not being connected in our visualization of people associated with the writing workshop. Of course, we know through others’ scholarship and accounts that these two did know one another but such information is not present in the description of the collections. Such, again, are the biases and gaps in archival data.
5. ^ In reflecting on the Belfast Group meetings, Simmons says, “I never got to know any of them well, perhaps because I was older” (Dugdale et al. 60). In The Ulster Renaissance, however, Heather Clark makes it clear that Simmons did indeed know Heaney, Longley, and Mahon well. Nevertheless he felt increasingly that he was “exclu[ded] from the poetry community” in Northern Ireland (88; see also 176–182 passim.).

doi:10.17613/4h79-m211 | PDF

As we worked on this project and looked at various iterations of the data, we noticed something troubling about some of the women we knew were associated with the Belfast Group: while they sometimes appeared central to the network at other times they were completely invisible. What was happening?

Women in the Group

While the members of the Group who ultimately became most famous were men, women participated throughout its nine years. According to the Group sheets held in Queen’s University Belfast and Emory’s Manuscript, Archives, and Rare Book Library (MARBL), Lynette M. Croskery,^[1] Marilyn Stronge, and Joan Watton (later Newmann) all participated during the Hobsbaum years, while Iris Bull presented poems during both the first and second phases of the workshop.^[2] In addition to these names from the archival record, Heather Clark writes in The Ulster Renaissance that, according to Philip Hobsbaum, the “founding members” of the Group included Hannah (née Kelly) Hobsbaum, Marie (née Devlin) Heaney, Edna (née Broderick) Longley, and Lucille Gregory (54-55). Group sheets do not exist for this group of women since, as Stephen Enniss notes in our overview to the Group, the pre-circulated sheets were only used for the poems being read in the first half of the evening. We know from recollections of participants, however, that Marie Heaney read her poems on one occasion and that Hannah Hobsbaum “wrote a play called When Rebecca Comes, which was workshopped in the Group” (Clark 59). Still, presenting one’s work to the Group was not a requirement for membership. As Clark writes, Philip Hobsbaum “often urged Edna to read her brilliant satirical poems at the Group meetings, but she always declined” (59). Indeed, she was a member of the Group before her husband-to-be, Michael, joined (see Clark 54–56).

Edna Longley was not the only woman who chose not to present—either formally or at all—at the workshop. This lack of participation is evidenced by Philip Hobsbaum’s statement to Clark in a 2000 interview that “It was hard to find women writers,” which suggests that he had made an effort to broaden the gender diversity of the workshop (59).^[3] But it was more than just “women" whom Philip Hobsbaum could not find; he went on to clarify that it was “even harder to find women writers who would have their poems eviscerated and excoriated by a group of their contemporaries” (qtd. in Clark 59). While women can flourish in an aggressive environment just as well as men, it’s also clear that Philip Hobsbaum was looking for people who would tolerate his way of running the workshop.^[4] Edna Longley essentially expressed this sentiment in a remark she made to a reporter for The Independent for a 1993 story they ran on Philip Hobsbaum; she described Group meetings as “very intimidating, run like a seminar in an autocratic way” (Ascherson, also qtd. in Clark 55). Seamus Heaney drove this point home by describing how “Philip concentrated on the poem sheet and hunched forward like a man on a Harley Davidson coming down the road at ninety” during the workshops (O’Driscoll 74-75). Those women who didn’t like to be “eviscerated and excoriated” by their peers—most of whom happened, by the way, to be men and whose ringleader created a “very intimidating” atmosphere—simply chose not to attend or present.^[5]

One gets an additional sense of the climate in Belfast’s literary circle toward women from the article that concludes the November/December 1970 issue of The Honest Ulsterman. Titled “Thoughts on Women,” it consists of a series of quotations from philosophers, poets, and essayists. The quotations chosen all address the subject at hand and present nothing short of misogyny. La Bruyère, Baudelaire, Blake, Montherlant, and Schopenhauer all make appearances. Yet as bigoted as their quotations are, they might be excused (inexcusably) as coming from “less enlightened times.” Not so the following two quotations, beside which the previous ones pale: from Patrick Kavanagh, “Silly feminists, who are never feminine, have created the notion that women like equality [...]. On a deeper level, there is even pleasure for a woman in the thought of being a slave”; and from Cyril Connolly, “A woman who will not feign submission can never make a man happy and so never be happy herself. There has never been a happy suffragette” (31, 32). Not only are these “thoughts” from contemporaries—Kavanagh having died only three years previously and Connolly still very much alive—but they are also from important voices on the literary scene. Kavanagh was an influential Irish poet whose work about the everyday influenced many who participated in Hobsbaum’s writing workshop, including Heaney and Longley, and Connolly was an equally influential critic and editor. The possibility that their statements might stand not just for themselves but also for the thoughts of the Belfast literary establishment is heightened by their publication in The Honest Ulsterman. James Simmons began the journal in May 1968, and it became “the most influential literary magazine in Belfast during the late sixties and early seventies” (Clark 86). According to Clark, it was “in reality, a mouthpiece for the Belfast Group workshop members, and gave local poets a collective identity” (97-98). Apparently this mouthpiece didn’t mind broadcasting terribly obnoxious ideas about women to its many readers. Since one of the editors of this issue, Michael Foley, has been described as “one of the most interesting satirists of this period,” there is a chance that “Thoughts on Women” was to be understood in such a vein (qtd. in Clark 100). Yet the fact that the article appears without comment by the editors (the other being Frank Ormsby) in either this issue or in the one that follows makes it difficult to see this as anything but rather oblique satire. In short, it appears that women hardly needed to attend Hobsbaum’s writing workshop to be eviscerated and excoriated.^[6]

In addition to the environment within the Group and Belfast itself, perhaps Edna Longley had another reason for why she chose not to bring her own poems to the workshop: she might have seen her role in the workshop differently, as a scholar and critic. At the time that Philip Hobsbaum began the Group, Edna, then, Broderick was his colleague in the English Department at Queen’s University Belfast; both of them were lecturers. While Hobsbaum moved on to Glasgow in 1966, Longley remained at Queen’s for the whole of her career, where she is Professor Emerita at the moment of our writing. Her scholarly publications—including books on Louis MacNeice, Yeats, and literature and revisionism in Ireland, as well as anthologies of, among others, James Simmons and Paul Durcan—show her interest in drawing attention to the poetry in Northern Ireland. She taught contemporary Irish and Northern Irish poetry in her classes, often that by her friends in the Belfast Group workshop (see Clark 34 n.99). Given her career, it seems possible that while Longley wrote the occasional poem, her interest in the writing workshop was chiefly scholarly, critical, and curatorial.^[7] Edna Longley’s work at the Group, in other words, was different but no less valuable. Indeed, Clark names Edna Longley and fellow critic Seamus Deane alongside Seamus Heaney, Derek Mahon, Michael Longley, James Simmons, and Paul Muldoon, as “writers who helped put Belfast on the literary map during the sixties and seventies" (6).

As with Edna Longley, it is important to consider the other types of roles that women played in the Belfast Group. In addition to penning and presenting her play, Hannah Hobsbaum helped to organize the weekly meetings; as Seamus Heaney wrote, Hobsbaum “and his wife Hannah kept open house for poetry” (Dugdale et al. 62, emphasis added). She also frequently acted as secretary for the Group, typing the poems and making duplicates for distribution. Rosemary Hobsbaum, Philip’s second wife, wrote in a letter to the authors that “Hannah deserves full credit for the help and encouragement she gave” during both the London and Belfast Groups. But the letter also points out that Hannah Hobsbaum acted as more than simply a secretary: “She was also a perceptive critic and occasional contributor of her own poetry.” Joan Watton Newmann makes a similar point, describing Hannah as “regal and there in her own right. Sure of what she needed to say, needed to write” (“Coming of Age” 118). Cilla Craig, the secretary in the Queen’s University Belfast English Department where Hobsbaum was a lecturer, also helped to type, copy, and distribute the poems (Enniss, “The Belfast Group Writing Workshop”). After Hobsbaum decamped for Glasgow, Marie Heaney worked with Seamus, as well as Michael Allen and Arthur Terry, to host and organize the Group meetings.

In short, while women did not become the most famous members of the Group nor were they its instigators, they played a critical role in making the Group happen.

Women in our Data

Although women such as Marie Heaney, Edna Longley, and Hannah Hobsbaum were central to the activity of the Belfast Group, we quickly discovered that they did not appear in our visualizations while many other women—even those who had not participated in the Group—did. Upon investigation, we have discovered two different reasons for this happening: archival bias and authorship.

While we have already discussed how our project depends on and is biased by archives, in the case of Edna Longley and Marie Heaney there is an additional wrinkle. When MARBL acquired the papers of Michael Longley and Seamus Heaney, included among those papers were some materials that concerned their wives. In the case of the Heaney papers, for example, there are twenty-six different photographs of Marie Heaney. She is accompanied by her husband in each of them, but she remains an important presence in this visual collection. Marie Heaney probably plays a role in the correspondence as well, although it is impossible to know since those materials are currently restricted from researchers. It stands to reason, however, that some of the correspondence is addressed to both her and Seamus. Yet the manner in which the material was catalogued—as the Seamus Heaney papers—effectively removes her as an addressee.

The difficulty of archival visibility is even more complicated in the case of Edna Longley. The Michael Longley papers include an entire section “by or about his wife” (Edna Longley Papers). Among the contents of this series are drafts and typescripts of the lectures and essays she produced in her academic career, as well as correspondence from the same time period. Yet despite her significant stature as a scholar—mentioned above and underscored in the series description, where she is described as an “eminent academic and critic”—her materials are subsumed within her husband’s papers instead of standing as a separate collection (Edna Longley Papers).^[8] Of course, there are reasons for this archival decision: Edna’s papers make up a relatively small portion of the Longley file; MARBL has a collecting focus on modern poetry rather than poetry scholarship; and including her papers with those of her husband might, paradoxically, make them more visible. Nevertheless, in both her case and that of Marie Heaney, the women are treated as appendages to the poets and are rendered invisible from the data as we collected it.

The other reason that these women did not initially appear in our data is due to authorship. Although both Edna Longley and Marie Heaney were founding members of the writing workshop, neither of them contributed Group sheets. So while they are both published authors and their works are found in MARBL collections, they did not author the particular object from and about which we collected information. Hannah Hobsbaum, on the other hand, did have her short play When Rebecca Comes workshopped at the Group and for that reason one would expect her to be included in the data. Surprisingly, however, that Group sheet is the only one of the ninety-five known to be extant that has no author noted on the sheet. Queen’s University’s catalogue record for the Group sheets (PDF link) had recorded the author as “[?].”^[9] Clark’s research helped us identify the true authorship of the sheet (59).

Since we felt that the contributions of these three women—among others—were critical for understanding the networks of the Belfast Group, we took steps to include them in our data. In the first place, we updated our version of the Queen’s University catalog so that When Rebecca Comes was attributed to Hannah Hobsbaum. Since that catalog record had been a PDF document, we had created an HTML version of it for the purpose of converting it into data. Adding Hobsbaum’s name and VIAF ID to the record was sufficient to have her appear as a new node on our visualization of Belfast Group authors by period, which shows authors connected to the different phases of the workshop. She is linked to Philip Hobsbaum in this graph not because she was his spouse but because he was the owner of the Group sheet in question. And since Philip Hobsbaum was only present for the first phase of the Group, we can infer that that is when Hobsbaum’s play was read; she is accordingly connected to the node that represents the Belfast Group from 1963–1966. As mentioned in the preceding paragraph, neither Edna Longley nor Marie Heaney show up on this visualization due to their not having authored Group sheets.

Our other network visualization simply shows individuals who were connected or affiliated with the Belfast Group in any way. None of these three women appeared in the initial versions of this graph at a one-degree connection to the Belfast Group. Both Edna Longley and Marie Heaney would have shown up in the two-degree version of the graph because they would have been connected to their husbands who were, in turn, connected to the Belfast Group. Given their roles in both founding, and in Marie Heaney’s case, helping organize the Group, this did not seem to be an accurate depiction of what we knew had happened. We wanted to affiliate them with the Group explicitly. What’s more, we wanted them to appear with Hobsbaum on the “bios” page of the site, which meant that they had to have a prose biography in our data.^[10] The approach we took allowed us to accomplish both goals at once. The biographies for many of the men featured on this site were drawn directly from their MARBL finding aids descriptions by targeting the biographical historical note (in the EAD element). Edna Longley has a brief biographical description in her series of the Michael Longley papers, but since it was not in the element, it could not get harvested appropriately. Our solution was to create a separate document for her brief biography and include it as a separate item to get processed in our data harvest. Neither Marie Heaney nor Hannah Hobsbaum had biographical descriptions as large as Edna Longley in their respective husband’s and ex-husband’s papers nor do they appear as individuals within Wikipedia, so we used other sources for their biographies. In the case of Marie Heaney, we used a series of brief biographies that were on the original Belfast Group site, again creating it as an individual HTML document to be included in the data preparation.^[11] A separate document was created for Hannah Hobsbaum’s profile, using the information about her in International Who’s Who in Poetry 2015. In all three of these documents we tagged the women with their stable Virtual International Authority File (VIAF) identifiers, which helped them remain distinct within our data set. Finally, in the case of Edna Longley and Marie Heaney we included lines in the code that explicitly declared them as “affiliated” with the Belfast Group. This step was necessary because the data process had previously defined affiliation with the workshop as authoring one or more Group sheets. (Because Hannah Hobsbaum did author one Group sheet, this step wasn’t necessary in her case.) Since both Edna Longley and Marie Heaney are named as founding members of the Group in so many different accounts, the affiliation seemed evident, even if we had to declare it ourselves in the document.^[12]

Once we added these three women to our data, it became possible to see them in our visualization of people associated with the Belfast Group. It is instructive to observe that both Marie Heaney and Edna Longley are tightly connected to the most active portion of the network. They, along with Hannah Hobsbaum, have a greater number of connections (degree) than many of the men in the network, and their influence in the network (measured by eigenvector centrality) is similarly larger than the majority of the other writers we are measuring. For instance, when looking at the GEXF data for individuals that have a direct connection to the Belfast Group, Edna Longley’s has an eigenvector centrality of 0.635, which makes her the eighth-highest ranked person in our data, with a score that is higher than both Ciaran Carson (0.514) and Philip Hobsbaum (0.451). By this one algorithmic measure, then, Edna Longley has more influence in this particular network than the person who founded the workshop. Marie Heaney (0.366) comes in just after Hobsbaum as someone who is more influential than more than half of the participants in the Group. When considering static representations of the full data set, as we used in our presentation at the 2013 Digital Humanities Conference, Marie Heaney and Edna Longley are positioned more centrally to the force-directed network than almost anyone else besides Seamus Heaney. (Hannah Hobsbaum was not yet included in our data and does not appear in the graph.)

Indeed, it was this positioning of the women near the middle of things, so to speak, that convinced us we should pay more attention to them within and around the Group.

Yet these positions in the network can be deceiving. For instance, eigenvector centrality, which reveals influence in a network, is determined not so much by the property of an individual but by how connected their connections are. In short, the eigenvector centrality score increases as the people one is connected to are connected to other people. In the case of Edna Longley, Marie Heaney, and Hannah Hobsbaum, their connections to their highly connected husbands boosts their scores significantly. Their marriages also account for the fact that they are placed more centrally in our early static graph. Since Michael Longley and Seamus Heaney are tightly connected and are connected to so many people, they remain close to the center of the network as they exert a sort of gravity on one another. Their wives are correspondingly pulled into their orbits, so to speak, in the representation. Even the simple count of their connections (degree) is skewed a bit because they have a connection to their husbands, which increases the degree count above the average member in the network who is not married to another participant and therefore only connected to the Belfast Group. While it must be pointed out that in this final measure the husbands get a reciprocal boost in their degree-count by virtue of their wives being part of the network, most ways of measuring the network show that it is the men who are affecting the placement of these women.

What these caveats mean, of course, is not that the women are necessarily less influential in the network than the men. Instead it points once more to the fact that our network reflects our data and that these data are clearly incomplete. For example, although Edna Longley’s nine connections place her in the upper range of people we measured in the Belfast Group, it is still a much smaller number than one might easily expect her to have formed with the members of the writing workshop. This would be even more true of Hannah Hobsbaum, who served as hostess for the first phase of the workshop, and Marie Heaney who helped play that role in the second phase. In particular, it’s worth noting that our data do not suggest any relationship between Edna and Marie Heaney. We know from Clark’s account of the Group that the Heaneys and the Longleys became good friends after meeting first at a workshop: “As the two couples grew closer, they often took drives through the County Down countryside in Heaney’s Volkswagen, singing Cole Porter songs” (56). It seems logical that Edna Longley and Marie Heaney should be connected in this case, but since the data draws almost entirely from their husbands’ papers, the relationship between their wives is left invisible.^[13] Equally absent is any indicator of a relationship that one might expect to see between Hannah Hobsbaum and either Edna or Marie. In an email to the authors, Hobsbaum makes it clear that she knew both of them: “Marie Heaney was always lovely, she and Seamus were a love match, she had the same standards and principles as myself. Edna also was very nice, I didn’t know her so well.” While we have Hannah Hobsbaum’s and Clark’s account, they fell outside of the material that we chose to include in our data set, and subsequently the women in the Group end up looking more isolated than they really would have been.

In a way, it might seem that this essay concludes similarly to those that have come before: with an acknowledgment that what we can observe about the Belfast Group clearly depends on our data and that we know those data are imperfect and biased. But the problems become more acute in the case of the women. Whether 1) women were unconsciously excluded as more regular participants, 2) women felt reluctant to participate due to the tenor of the workshop or the culture in Belfast at the time, or 3) if “there happened to be fewer women who were interested,” as Joan Watton Newmann wrote in an email, the contemporary practices of archives and literary historians contribute to their erasure. Our intervention cannot fully explain what roles Hannah Hobsbaum, Edna Longley, and Marie Heaney played nor the relationships that they formed with the other members of the Group, but we hope that restoring them to the network makes for a better—if only marginally so—picture of the networks of poetry that flourished in 1960s Belfast.

This essay was peer reviewed by Geraldine Higgins and Nathan Suhr-Sytsma.

1. ^ In The Ulster Renaissance, Heather Clark writes that “later entrants” to the Group, according to a 2000 interview with Hobsbaum, included “Lynette McCroskery” (55). Stephen Enniss also provided us with a 1998 email from Hobsbaum in which the latter mentions “McCroskery” (Hobsbaum, Email). The spelling that we use on this site—Lynette M. Croskery—originates with the record of materials at Queen’s University Belfast (PDF) and appears on the Group sheet itself. Unfortunately, the only mentions that we have been able to find of this author of three short stories lead back either to the Queen’s University records or to Hobsbaum’s statements. We suspect that “McCroskery” is the right spelling and that the spelling on the Group sheets is a bad transcription of “M’Croskery,” as the patronymic is sometimes abbreviated. Yet since our data are constructed in part on the basis of the records at Queen’s, we have chosen to use that spelling. While the difference of spelling is slight in either case, it is far more important to observe again how a woman in the network around the Belfast Group falls through the cracks of scholarship and editing.
2. ^ One of the difficulties of accounting for the women in the Group is the fact that many of them participated under different names than they would later publish under. Of special interest here is Joan Watton Newmann; she participated in the Group as Joan Watton, which is the name on all her Group sheets. When she married, she took the name Newmann, and it is under this name that she has published all of her work, as well as run the Summer Palace Press with her daughter, Kate. In deference to the archival record, we have kept the name “Joan Watton” on the Group sheets, but refer to her elsewhere on the site as “Joan Watton Newmann” for clarity.
3. ^ In an email to the authors, Joan Watton Newmann wrote, “It was a smallish group and at no time suggested that there was an absence of women, whether purposefully or accidentally. Philip gathered writers and we were a disparate lot. There happened to be fewer women who were interested.”
4. ^ It is worth reiterating Joan Watton Newmann’s perspective, quoted in our first essay, that “the ethos [Philip Hobsbaum] created and sustained was one of pleasure and discovery. There was no space for destruction” (Newmann, “Coming of Age” 118).
5. ^ In a case of history repeating itself, Hobsbaum’s shortage of women participants in the Belfast Group mirrors that from his previous writing workshop in London, which is known simply as the Group. Clark records Hobsbaum’s comment that “We were very short of women in the Group” (49). One reason that they were short of women is that Hobsbaum refused to let just anyone in and markedly denied admission to Sylvia Plath while allowing her husband, Ted Hughes, to participate. As Clark writes, “Plath’s rejection went unnoticed in the 1963 Group Anthology, edited by Hobsbaum and [Edward] Lucie-Smith, who, in their foreword, claimed that ‘anyone who asked if he could come was welcome to do so. No one has ever been expelled or excluded.’ Hobsbaum’s use of the male pronoun here hints at the social barriers which would discourage women writers from joining Hobsbaum’s London and Belfast Groups” (49). In his 2000 interview with Clark, Hobsbaum admitted, about Plath, “I was wrong of course” (49).
6. ^ It is also important to note that the composition of the Belfast Group did not change significantly after the departure of Hobsbaum. In its second phase, two of the most significant additions to the workshop were Paul Muldoon and Ciaran Carson, who were about a decade younger than Heaney and still in school. Muldoon, in particular, studied with Heaney at Queen’s. Yet another of Heaney’s students, Medbh McGuckian, never became a member of the Group. McGuckian who was born in 1950 is a contemporary of Carson (b. 1948) and Muldoon (b. 1951) and like all three of them was born to Catholic parents. There is only a single account of her having attended the Group during the time that the Heaneys were its organizers. In a conversation with Nuala Ní Dhomhnaill, McGuckian mentions that in her final year at Queen’s in 1972, Heaney taught one of her seminars: “He was the first person who didn’t make me feel that poetry was a closed shop. I got up the courage to say that I would like to be a poet, and although I hadn’t yet put pen to paper he invited me to the group, and Paul Muldoon was there. There was this openness and friendliness that I trusted” (592). This invitation must have been extended in the second phase of the Group when Seamus Heaney was the chief organizer. His recollections, however, are less than certain about who attended at that point: “Did Jimmy Simmons attend at that stage? Harry Chambers? I’m not sure. I remember Frank Ormsby and Michael Foley and Paul Muldoon showing up, possibly Ciaran Carson and Medbh McGuckian, but I’m not at all clear about the who and the what of it” (O’Driscoll 106). Given the facts that McGuckian’s attendance is uncertain even for Heaney and that she is not mentioned in any of the other recollections of the Group nor by scholars who have written about it, one might assume that she only attended the one time, much like Derek Mahon earlier. Perhaps McGuckian’s absence accounts for Seamus Heaney’s statement in The Honest Ulsterman’s “Symposium” that “When the second act opened in my own house [...] some of the old characters had departed [...] and a crowd of gifted boy actors were in the wings to claim the stage” (Dugdale et al. 63, emphasis added). Or perhaps this statement explains why McGuckian did not make an appearance.
7. ^ Clark reports that Mahon “was more unnerved by [Edna’s] frank appraisals [of his poetry] than he was by her husband’s” (35).
8. ^ In the case of Marie Heaney, we only suspect that some of the correspondence addressed to Seamus is also addressed to her. In the Longley papers, we know for a fact that much of the correspondence that is catalogued as addressed to Michael is fact to both him and Edna. Yet these materials are only connected to him in the finding aid. The result is that in our data model, the connections that these letters represent only get made to Michael.
9. ^ While the Group sheet of When Rebecca Comes at Queen’s University has no author on the sheet, a handwritten list of contents in the collection has been amended to indicate that Hannah Hobsbaum is the author. It is unclear when the addition was made, why the catalogue record has not been updated, and, most perplexingly, why the author was not listed on the Group sheet in the first place. As Stephen Enniss notes in our overview of the workshop, Hannah was one of two typists for the Group sheets during the Hobsbaum phase. Perhaps, we thought, in this case she acted as the typist and left her name off out of modesty. Such a theory might account for the fact that her play is the only known Group sheet that is anonymously authored. When we asked Hannah Hobsbaum about this possibility via email, she responded, “I had hoped that When Rebecca Comes would have gone into oblivion.”
10. ^ We set these minimum criteria for having a profile page to ensure that we did not display empty pages. Since the descriptions in our data are pulled from Wikipedia and the archival collections this effectively excludes anyone who is not notable enough to be in Wikipedia or have an archival collection.
11. ^ The biographies from the original site for the project also included a description of Arthur Terry. Without that statement, he would not have appeared in the site’s “bios” page either since MARBL does not have his papers. The fact that Terry would have been just as invisible as these women suggests that gender is one factor, but not the only one, in practices of cataloging and literary history that ultimately determine who is visible and who is not. The exclusion of both Terry and Edna Longley from the archives, in one form or another, also points to the preference of archives for poets or creators over scholars.
12. ^ By way of clarification, in the early process of enhancing finding aids for our data harvest, we updated the Medbh McGuckian record because we knew she was from Northern Ireland and was connected to the major players in the Belfast Group. Ultimately, her finding aid is not explicitly included in the data set because her attendance alone does not seem enough to warrant her explicit affiliation as part of the Group. However, McGuckian’s finding aid is listed as a related material on the Seamus Heaney, Michael Longley, and Ciaran Carson finding aids, among others, and one of the scripts from the data harvest process instructs it to scrape related materials. When it does so, it does not associate her with the workshop but rather adds additional context for the second-degree view of the visualization of people associated with the Belfast Group.
13. ^ Marie Heaney is connected with Michael Longley as a correspondent, since she appears in the list of selected correspondents in his finding aid. This letter is addressed to Michael Longley, thanking him for two poems he sent her on the occasion of the birth of her daughter, Catherine. In this letter’s case, Edna is not an addressee, which would seem to corroborate the lack of connection between the two. But Marie asks Michael to “give my warmest regards to Edna” and occasionally addresses the Longleys in the plural in the body of her letter. Both of these suggest that a relationship that does not appear in our data.