[Coral-List] push for more reliable research in ecology

[David Blakeway]

Thanks Doug.

I’m aware that some of the things I’ve said may be offensive. I apologise
for that. They were not intended as an attack on science or scientists. I
know that a LOT of thought and effort goes into designing surveys and
experiments and building ideas, not to mention the teaching, supervising,
administration etc. It’s a tough, important, job.

But science is not about reality, it’s about creating representations of
reality, and that’s always going to come with inherent risks of falsity.
Scientific progress is never an incremental series of linear steps, it’s
all over the place. It’s easy to get waylaid by dangerous ideas that are
compelling, and backed by apparent evidence, but false. These ideas have
negative quality; they can set a field back for decades.

I’ll give an example from my field of reef geomorphology. The details are
arcane and would take time to explain, so I’ll shortcut: when reefs were
first investigated (by white Europeans) in the early nineteenth century,
their forms were attributed to primarily to process A, which is a
reasonable proposition and one which I believe is correct. Late in the 19th
century, information came to light which made process B a distinct
possibility. Process B was proposed in the 1940s and elaborated in the
1950s and 1970s. When the issue was tested by reef coring in the 1980s and
90s, process B was falsified. Since then, articles supporting either A or B
have been regularly published, the latest supporting B in 2020. What is
happening? I’m really not sure, but I believe that, if I presented some
reef images to you and asked you how those patterns formed, process A would
immediately come to mind.

The sad thing about all this is that an article published in 1918 had
already addressed the issue and concluded (paraphrasing) ‘these forms look
like process B but are more likely to be process A, for reason c. Reason c
being a very excellent reason.

If the simple logic of the 1918 message had been taken to heart, we never
would have had to bumble through this whole (ongoing) mess. This is
science, and we all have to accept that it could be our science, especially
if we haven’t read any 19th and 20th century literature lately. Admittedly,
that 1918 article was in a very obscure journal, but it was there. More
importantly, the logic embodied in the article was always there in the
reefs.



Do I think more than half coral reef science is false? – well I cringe from
that thought because it seems insulting to reef scientists. But that
feeling is about people, not ideas. The ideas have to be up for
investigation. So I would answer ‘I don’t know. Which ideas are critical?
Have they been adequately tested?’ Yes, synchronous spawning occurs and yes
there are craters on the moon. And when highly competent people build
logical ideas based on those facts, using explicit methodology, and the
ideas withstand testing, I believe them. But they are still just ideas. I’m
grateful when people point out my mistakes. Even though it’s a bitter pill,
it’s much better than continuing down the wrong road. And if the
interaction is done right, it can be wonderful. Don Kinsey told me that in
his early career a competitor emerged, Steve Smith. But instead of
competing they joined forces, produced some pretty awesome science, and had
a good time doing it. I really dislike seeing nasty comments and rebuttals.
It’s unseemly and can lead to long-term harm, especially to students in the
respective labs who may have had nothing to do with the original argument.
Those authors should speak to each other before the comment appears out of
the blue. As Doug & I did, now we are comrades. With respect to this issue,
I have wondered whether I should contact André Droxler about our ‘Darwin
was wrong’ thread. We (especially I) were quite critical behind his back.
The thing is, I’m pretty sure if I met André we’d get along great, have a
beer and a snorkel, err I mean a snorkel and a beer. Maybe I should do it.
It’s pretty crazy when people end up disliking each other just because they
share a common interest.

Conspiracy theories are also ideas, up for testing. The conspiracy theory
that the earth was created in seven days I don’t spend much time on. The
conspiracy theory that Facebook is selling data to the highest bidder, and
those data may be used to identify and track people that might rebel
against a regime, well I don’t spend much time on that either, but I think
it’s better than the seven-day thing.



I share your concern about the credibility of science being damaged by
trumped up claims of falsity. Ioannidis’ article was actually criticised
far more for its attention-seeking title than its content. And ‘Darwin was
wrong’ had become a constant internet meme well before the Droxler & Jorry
article. Although virtually none of the ‘Darwin was wrong’ claims have any
substance, they still have the potential to discredit science, a point made
by Bob Buddemeir early in that thread. However, these are shallow claims
and easily dismissed, in theory at least. I think there may be a greater
risk to our credibility, which I’ll outline below.



Thanks for digging up the data on numbers of scientists. That 0.23 does
seem too low. It certainly won’t get you tenure anywhere. Even I can hit
that target. Say there’s 200 scientists in a research field and each writes
three papers per year (we can forget the multi-author aspect as it cancels
out). Add some papers from your other fields of interest, some stats
papers, some technical documents, and you already need to read two papers
per day to keep up. That’s difficult, especially when there’s little direct
gain from reading a paper compared to authoring one. In fact I think
there’s very little reading happening. I see plenty of subtle clues
supporting this, and one not-so-subtle clue: I don’t even read many myself!
I’ve downloaded hundreds of papers, because they interest me, and then
haven’t read them!



I want to return to the Hirsch 2005 paper here, because I don’t think that
one’s being read either. Virtually everything I see about it, praise and
criticism, is focused entirely on the sweet little nutshell, the index
itself (the concept ripped from Eddington). The Wikipedia h-index page, for
example, lists eight criticisms but they are all derived *from* the index,
there are no criticisms about the derivation *of* the index. Now, we can
argue about the importance of the index, and we have, but just the fact
that it’s in everyday use among scientists means that it has become *some*
sort of credible scientific icon. And, therefore, that if it were shown to
have no logical basis, well, I reckon we have our pants down.

In fact, what I believe, and this needs to be checked by someone better at
logic than me, is that it is a complete logical mess: it bootstraps it’s
own axiom from circular reasoning applied to a hopelessly biased sample.
And is totally vulnerable to attack if someone outside academia starts
investigating. Not a malevolent person, just a *seeker*. I’d like to meet
the seeker, but I think we should be doing the job ourselves.

I’ll look dumb if I’m wrong of course, but I don’t mind that, because I
think there’s significant credibility at stake if I’m right. And I think
this whole issue is very symptomatic. Nobody’s investigating axioms. We’re
all expounding on what we think, not investigating why we think what we
think, not looking back, just not being good scientists.

The seeker would be a f*****g good scientist.