Fake Experts

You know who they are – those organizations that have words like “freedom” and “rights” “choice” and “consumer” in their names but always shill for corporate interests…those occasional MDs or engineers creationists find that will say evolution has nothing to do with science. They are the fake experts.

But how do we tell which experts are fake and which are real?

To figure out who is a fake expert you have to figure out what a real expert is. My definition would be a real expert is someone with a thorough understanding of the field they are discussing, who accurately represents the scientific literature and the state of understanding of the scientific enterprise. There has been some other other discussion on scienceblogs from Janet at Adventures in Ethics and Science, it also reiterates some of the same points in relation to what she feels comfortable discussing as an expert. It also stresses the importance of context in evaluating the validity of expert opinion. But I’m not the god of the dictionary so let’s consider some other definitions.

The OED gives the definition simply as “One whose special knowledge or skill causes him to be regarded as an authority; a specialist. Also attrib., as in expert evidence, witness, etc.”

I don’t think this is adequate to describe what we really mean though, that is, how do you identify a trusted source of scientific information?

Legally (in the US), scientific expertise had been defined by whether the testimony the expert provided conforms to the so-called Frye rule from 1923 until 1993 when the Daubert vs. Merrel Dow Pharmaceuticals case changed the definition to be consistent with the federal rules of evidence. The Frye rule was that scientific testimony was valid if the theory it was based on was “generally accepted”, that is it was admissible if the theory on which the evidence was based had a somewhat arbitrary critical mass of followers in the scientific field.

In many ways Daubert was a big improvement, although it puts more onus on the judge to determine if the science presented should be considered valid as it merely stated that experts were defined by the federal rules of evidence which allow the judge to determine:

If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise.

(A good article on this issue here from the NEJM and a more updated article.)

Luckily the justices didn’t just leave it at the federal rules of evidence and Blackmun created a set of guidelines for judges to determine if the expert was “reliable”. They require the theory presented by the witness to have undergone peer review, show falsifiability, empirical testing, reproducibility, and a known error rate for a scientific theory to have some validity in addition to the general acceptance rule of Frye. While the individual states remain a patchwork of Frye, Daubert, and Frye-plus rules for admissibility of evidence, at least federally this is the new requirement (although it still does suffer from being a bit vague).

The experts that present such evidence must have some credentials and/or experience with the discipline, and the evidence they present must pass these tests. It’s actually not a half-bad way to identify a trusted source, in particular if the judge is intellectually honest about the witness meeting these requirements. Although the law currently allows a lot of latitude on this as it’s really up to the judge to determine if the expert testimony satisfies the Daubert requirements.

The commonalities between the different accepted definitions are that experts have experience in their field, and they can provide answers that are consistent with the state of knowledge in that field that are useful. The legal definition appears more stringent, in that it requires the expert to speak in a clear fashion and discuss science that actually meets Popperian requirements of epistemology(falsifiability, testing, etc.) – but I’m not about to jump into that quagmire today.

Clearly, the exact definition of what an “expert” is still eludes us, but it becomes readily apparent from the legal, dictionary and common practice definitions employed by scientists what experts are not. They aren’t merely an empty set of credentials and they aren’t merely people who have at some point published in some random field. Even the rather silly expert wiki would seem to agree on this.

Therefore I would say a fake expert is usually somebody who is relied upon for their credentials rather than any real experience in the field at issue, who will promote arguments that are inconsistent with the literature, aren’t generally accepted by those who study the field in question, and/or whose theories aren’t consistent with established epistemological requirements for scientific inquiry. Sheesh. I just described Michael Egnor, Bill Dembski, Michael Fumento, Patrick Michaels, Steven Milloy, Richard Lindzen…

So, in honor of the false experts hired by everyone from creationists to global warming deniers, I present to you, the thinking chimp. Our mascot of the false expert, who isn’t as good at telling you accurate information about science as he is at flinging poo.
i-489dd819efedba2ae35c8ed120ac2485-3.gif

**Janet points us to another post of hers discussing how to identify a trusted source.

Comments

  1. #1 Scott Belyea
    May 2, 2007

    I suggest that there’s an additional characteristic of a responsible expert – that the expert is aware of and open about the limits of their expertise.

  2. #2 bob koepp
    May 2, 2007

    the “and/or whose theories aren’t consistent with established epistemological requirements for scientific inquiry” creates an opening too large to be called a mere loophole. Between various flavors of positivism, Popper, Feyerabend, Kuhn and assorted other non-experts, you won’t find a consistent account of the epistemology of science.

  3. #3 MarkH
    May 2, 2007

    Well, I think Blackmun chose a Popperian flavor for his definition, and that tends to be what I identify with, but you’re right, there’s a lot of wiggle room there.

    This is why the internet hive mind is helpful. While I dread a discussion of epistemology (I’m no expert) as they can really get bogged down with a lot of lingo and endless disagreement, I would like a general idea if people could agree on a few elements of each epistemological theory that would help identify someone who has a good idea of how science works. If we can identify the requirements that are common to each philosophy, it can go a long way to figuring out what people can agree on is consistent with a trusted source of information.

    Also, as a side note, I also like how Popper is inconsistent with more of the creationist arguments, as they love their “rigorous inductive argument”.

  4. #4 Stuart Coleman
    May 2, 2007

    “Our mascot of the false expert, who isn’t as good at telling you accurate information about science as he is at flinging poo.”

    I think that’s an understatement. False experts are terribly good at flinging poo, to the point where they can explain the nuanced techniques for poo-flinging at maximum efficacy. They might not be experts at one they pretend to be, but they sure are experts at throwing excrement around.

  5. #5 Torbjörn Larsson
    May 2, 2007

    If we can identify the requirements that are common to each philosophy, it can go a long way to figuring out what people can agree on is consistent with a trusted source of information.

    I think there are separate issues here, to recognize the methods and to explain them. The first is sufficient here, and is incidentally easier.

    For example, testability of some predictions of a theory is simply necessary to be able to eventually reject those that are in error. But we could discuss at length exactly why testing works, exactly how to do tests and what modifying a theory really means.

    I think it is in these situations that testability (“Popperianism”) is a simple and powerful requirement. It draws out exactly the points that denialists wants to gloss over, observability (of facts) and predictability (of theory). For example, the assumed ‘design’/’information’/’complexity’ of creationist denialists that can’t be defined so we can measure it, or the unspecified designer of ID denialists that can’t be described so we can predict what it will do.

    Usually practicing scientists seems to be much more satisfied with demanding testability (for example inflation and string theory) than philosophers are. I wonder why that is.

  6. #6 G Felis
    May 2, 2007

    Torbjorn: Philosophers of science are perfectly comfortable with demanding testability of scientific theories, but in terms of a formal logic, Popper’s falsifiability requirement as originally formulated simply does not work. That does not mean that philosophers of science have rejected Popper – except in this specific, limited sense. Nor does it mean that philosophers of science reject testability as fundamental to defining/understanding/practicing science, although casual readers (or willful deceivers) often take it as such.

    This is just one of those areas where a little knowledge is a dangerous thing. Philosophers pointed out a flaw in the formal deductive reasoning Popper offered for falsificationism (an error which Popper later acknowledged). Later, non-philosophers pick up some half-baked, trickled-down, game of telephone version of that criticism and say, “Popper was wrong! Falsificationism doesn’t work. So there!” But it ain’t necessarily so.

  7. #7 MarkH
    May 2, 2007

    Hmm, I was hoping to hear some defenders of induction and Kuhn to balance things out. Instead you guys are all supporting Popper!

    Oh well. I’m fine with using Popperian requirements as part of the definition. So instead of saying that they have to agree generally with scientific epistemology, let’s say they have to understand the importance of testing, falsifiability and predictive power of their theory.

  8. #8 jtdub
    May 2, 2007

    Can anyone recommend a good book in this philosophy of science vein? Epistemological concerns seem to be coming up more and more often on ScienceBlogs, reminding me of my poor background in this area.

    Maybe that’s because I started out with Kant. Ack.

  9. #9 bob koepp
    May 2, 2007

    jtdub – Kant, eh? You might want to look at Hume, who Kant credited with waking him from dogmatic slumber. Or, moving much nearer to what is now called philosophy of science, Peirce asks all the right questions. Instead of recommending any particular book, may I suggest that you visit the online Stanford Encyclopedia of Philosophy — search “philosphy of science,” browse, and take advantage of the bibiographies

  10. #10 Torbjörn Larsson
    May 3, 2007

    Philosophers of science are perfectly comfortable with demanding testability of scientific theories, but in terms of a formal logic, Popper’s falsifiability requirement as originally formulated simply does not work.

    Thanks for the information. That wasn’t the impression I get elsewhere. But it is admittedly difficult to understand what the current picture is.

    Since I haven’t studied Popper and his detractors, I don’t know about the specific arguments. If you can share them here, please do. As I understand it Popper has described how testing works, and indeed a theory can be tested and then it, or its status, is changed. The problem may be that he wants to “demarcate”, to identify scientific theories.

    Often it seems to me the philosophical analysis doesn’t study capable models of the factual situation. This seems to be such a case; theories works, testability is deemed important by the practitioners, falsifiability seems to work.

    What I have seen is that the criticism is along two similar paths. (And yes, I use Stanford’s encyclopedia, it is a great resource. I just don’t think the epistemology question is about science methods.)

    First, that there “are no critical tests” and that “theories can be saved”. But tests changes which theories we can contemplate, changes their form, constrains their parameters, or otherwise changes our trust of them. (Keeping false but useful theories comes about for other reasons too, so it isn’t a specific problem for falsification.) Testing is powerful.

    Second, philosophers criticize the use of falsification to “demarcate” or uniquely identify which is a scientific theory or not.

    I’m not sure how these criticisms are described by formal logic. But specifically here, they don’t bear on the problems with denialists ideas. If falsification can be less efficient or happens to include ideas which is not science are separate problems. Here we need to know that lack of falsification means it can’t be science.

  11. #11 jtdub
    May 3, 2007

    Wow. That is a great site- thanks bob!

  12. #12 Torbjörn Larsson, OM
    May 3, 2007

    “I use Stanford’s encyclopedia”

    Specifically, I used their installment on Popper above.

  13. #14 Oldfart
    May 19, 2008

    Thanks for the Popper turn-on.
    The problem I see with Popper (from my minuscule reading of him) is not the falsifiable problem but the other side of his ideas. The fact that the scientific method cannot lead to certain proof of a theory. Therein lies a crack that becomes a gaping fault into which creationism and ID and other woo festers either by pointing out that evolution cannot be proven or by stating a particular version of woo “could” be true.

    Does he actually state, or has anyone argued that so-called “open” statements cannot be science? Many statements made by real scientists are not falsifiable.

The site is currently under maintenance and will be back shortly. New comments have been disabled during this time, please check back soon.