growth of paleo - BFV results

From
Date
Reply-To
Here is my response to Norm's queries on my literature study.  
Basically, I agree that the conclusions may not apply to the entire  
discipline of paleo but my reasoning is different. In order to back up  
the results I present a largely independent study of the Bibliography  
of Fossil Vertebrates (BFV) after my response.

1) The database is essentially complete up to a little past 1990 (the  
year covered by the last issue of BFV). All paleo journals known to me  
were checked, including such journals as PALAIOS, Historical Biology,  
and even "really" foreign journals such as Vertebrata PalAsiatica that  
publish little or nothing on North American mammals. The database has  
taken eight years to assemble and I have made every effort to complete  
it for the period from 1960 on.

2) I believe that journal articles are not getting longer. If  
anything, the average article length in my database should decrease  
through time because I was much more likely to add an abstract to the  
database if it was recent. For example, I went through the entire last  
10 years of the Proceedings of the Nebraska Academy of Science looking  
for suitable abstracts. The only GSA and SVP abstracts used in the  
database are post-1980.

3) I didn't say mammal paleo was declining, just that the field's  
growth had come to an abrupt and total halt. The number of papers/year  
is about the same now as it was ten years ago (actually less in VP as  
a whole, see following).

4) If the "problem" is an increase in journal backlogs, that has  
nothing to do with the real problem - a decrease publications in the  
field, which seems to me the best measure of overall research  
activity. I also doubt that any amount of journal backlogging could  
keep the curve flat for an entire ten years if the number of  
acceptable submitted manuscripts had actually increased at 7% per year  
(and therefore doubled) during that interval.

5) All that said, the database pertains only to _North American fossil  
mammals_. Maybe this area has become a paleontological backwater over  
the last ten years or so - perhaps many North American mammal  
paleontologists have shifted their research to African, Asian, South  
American, and Australian fossils, so there are more North American  
workers but they are working less often on North America. In order to  
check this I made a quick search of recent Bibliography of Fossil  
Vertebrates issues, counting pages in the "author catalogue" section,  
with the following results:

Volume	Pages	Refs.	Ratio
73-77	52*	1214*	---
1978	60	1336	1.10
1979	83	1885	1.41
1980	97	2042	1.08
1981	172	2179**	1.07
1982	190	2407**	1.10
1983	210	2661**	1.11
1984	204	2585**	0.97
1985	179	2268**	0.88
1986	163	2065**	0.91
1987	156	1977**	0.96
1988	148	1875**	0.95
1989	159	2015**	1.07

* average for five years (260/5 and 6068/5)
** estimated using 12.67 references/page

I don't have 1990 on hand, I'll have to hike over to the library to  
check it.

The estimates require some explanation. Up until 1980 the bibliography  
numbers references consecutively, so you can easily look up the total  
number. After that I had to use an estimate. The font size and type  
changes in 1981 (bigger) and again in 1985 (smaller). For the four  
pre-1981 volumes references/page comes out to 23.3, 22.3, 22.7, and  
21.1, which is reasonably consistent. So I estimated the number of  
references/year by multiplying across using a count of the average  
number of references on pages 30, 40, and 50 of the 1984 and 1989  
volumes for the respective intervals. The resulting correction factors  
are 14+13+12/3 = 13.0 and 12+12+13/3 = 12.33. Surprisingly, the  
mid-80's font change appears to make no significant difference in  
number of references/page, so I used a grand average correction factor  
of 39+37/6 = 12.67 for 1981-1989.

These data show a _decrease_ in publications starting after 1983. They  
also show a rapid exponential increase before that, on the order of  
7-11% per year (see ratios; 1978-79 increase is anomalous, perhaps  
related to changes in BFV reference collection methods or budges). The  
results accord well with my earlier conclusions, and if anything are  
even more disturbing. It would seem impossible to argue that North  
American workers (or mammal workers) are simply working on different  
continents or on different groups. Of course, the data are extremely  
rough. For example, the correction factor could be more precise and  
there is some "slop" in each bibliography with older references  
cropping up after being missed in earlier volumes. However, the BFV is  
extraordinarily well-researched and I find it very hard to believe  
that it routinely misses the 5 or 10% of the literature per year but  
not _every_ year - wildly varying inconsistency of this sort would be  
needed to corrupt the data enough to affect the results.

Does someone at Berkeley have easy access to actual counts of  
references/year based on the BFV? Or are we stuck with estimates like  
the ones given above? Does anyone else have data like these, perhaps  
for micropaleo, macroinverts, or paleobotany?