More on a model of publication numbers, June 2010
This adds something to ‘My errors and some numbers too’, posted on this website.
I did some work in 2010 to correct figures for the overall volume of poetry publications in the period 1960-97, offered on this website. The work was based on spreadsheets with long lists of names. There are three values which the spreadsheet exercise is there to correct. One of these, the curve of the male/female ratio, is not controversial, fairly secure, and does not require exhausting scrutiny. The other two are less secure. They are the figure of circa 7000 for poets publishing in the period, and the count of roughly 2000 underground poets within that figure. I have been doing some work to check these and confidence is hard to obtain.
I discovered that for three years in the 1970s there was an annual issue of Poet’s Yearbook, a publication which on the basis of returns from over 900 publishers issued a list of all poetry publications for that year (from June to June). The first result from this was that the annual listings produced by the Poetry Book Society were hopelessly incomplete.
Using this excellent new source (PY), I typed up lists of poets publishing in successive years, 1976 and 1977, and by matching these derived a count of the overlap between the two years. 106 poets recurred between 1976 and 1977, out of 628 in the first year. So in theory the whole list would recur over 6 years. So a count of the total pool at that point is 6x628 which is 3768, less the 628 double appearances, so 3050. This would be the ‘model’ total pool of poets active in that time. Assumptions about the length of a working life would allow us to scale this up to the whole 40 year period - e.g. double 3050 or 2.4 times 3050 (depending on the assumption used).
Some poets born in the 1880s were still publishing in 1960 and 1961. Some poets were publishing for the very first time in 1997. We have to consider poets entering and leaving the pool in order to get at the count for all poets active in the period 1960-97. Assume the pool arrives in 40 exactly equal annual cohorts and each cohort leaves after exactly 40 years. This means that the pool in 1977 includes exactly 20 cohorts and therefore the set of poets active in the whole period is double the pool active in 1977. Therefore this count for the whole period is 3050x2, equals 6100.
This is an idealisation because the cohort entering in 1960 was certainly much smaller than the cohort entering in 1975. On balance this means 6100 is an overestimate and so the total in this model would be less than 6100, perhaps between 5000 and 5500. The 40 cohorts are all of different sizes and they probably increased in steps from 1960 onward. If we adjust the multiplier to 1.6x 3050 we get a count of 4875 poets.
This model is accurate to within half an order of magnitude (he said modestly). Its real value is to get a ‘fix' on the other model, where we estimated 33,000 books published and using a bludgeon translated that into 7000 poets. The two figures critique each other and give us a hint of where the true figure must lie.
The PY lists also give us information, only at a point in time but quite thorough for that point, of the balance between male and female poets. Counting entries in Poet’s Yearbook 1978 (for publications between June 1976 and June 1977) we find:
71.9 % male
indeterminate by name 4.9%
I don’t have comparable counts for the 1950s or 1960s, but it is clear that the scene was male-dominated in 1976. Using counts from selective sources like anthologies and the lists in British Council pamphlets, it is possible to guess that female participation in the 1950s was around 10 to 15%, and so we can suggest that this share was growing up to 1976, in line with greater access to higher education of female students, and a reduction in the rigidity of gender roles.
One thing that PY yields is a critical comparison with the lists in the Poetry Book Society lists for each year. In 1974 the PBS lists 450 books + 61 anthologies, in 1975-6 they list 859 titles. This figure is identical to the one in Poet's Yearbook so the jump from 1974 to 1975 is probably due to copying the figures from PY! It follows that the count in the PBS list is probably far too small for the entire series. The 'hike' blows their credibility. So any figures based on their count for the period 1960-75 are in doubt - as too low. My guess is they ignored little presses unless forced to include them. (Note that the PY year runs from June to June.)
It is possible that 74 to 75 was a growth year, but post the ‘oil price shock’ inflation had already taken off and this does not seem like the basis for sharp growth. So a jump from 450 to 850 titles in a single year is due probably to a better means of collecting information.
Why was I interested in these numbers? It has to do with the completeness of 'Affluence' (the overall project which includes all my books on British poetry). I covered 140 poets from the period in 'Affluence'. Selectivity was a big issue for a lot of readers. The numbers let me get at selectivity - and the answer is that everything is drowned in it. 140 poets is just a drop in the ocean. Almost everything is forgotten. Another answer is that "in all this warehouse of dead print, there are a number of poets who really count, and the cognoscenti know who they are“. So by missing out some of those poets I would be committing errors. But either the cognoscenti don't cover the terrain or they keep their knowledge to themselves.
I don’t think anyone would go and read all 850 volumes published in 1975-6. At some level, we all agree on one basic thing - that most of the poets publishing were wasting their time. Quite a few of us share the same question: how do you know which of the 850 books are worth reading?
As an aside, Poet’s Yearbook was only published for three years. It is a fabulous source but there was really no market for it. What do you with a list of 850 poetry books? ST Gardiner edited it and did all the work. It is coincidental that they cover a lot of the period in which Poetry Review stopped running reviews and so drops out as a reference source. PY is a high quality publication but it only lasted for three years and unfortunately its figures cast doubt on the other series we have, which run for longer. This whole area is paved with uncertainty.