• Tuesday, May 29, 2012

Previous

Next

Study of Fish Suggests the Value of Uninformed Voters

December 15, 2011, 2:12 pm

 

Could the golden shiner prove the salvation of our democracy? New research involving the common minnow suggests that uninformed voters are important in blunting the effects of political extremists. Getty Images

As Congress proves itself increasingly dysfunctional and captive to extremists, lots of people may be asking themselves: What kind of fish-brained voters keep electing these guys?

A team of researchers led by a Princeton University biologist has now studied that question and concluded that without all our know-nothing fellow citizens, things might be even worse.

The team, led by Iain D. Couzin, an assistant professor of ecology and evolutionary biology at Princeton, carried out its work with a type of fish known as golden shiners. The group trained some of the fish to associate food with a blue target and trained a smaller number of the fish to associate food with yellow, a color the fish more naturally prefer.

Placed together, most fish pursued yellow targets, suggesting the smaller group’s more intense desire for yellow overwhelmed the larger group’s numerical advantage, Mr. Couzin reported. But as fish without any training were added, the group increasingly favored the blue target, he said.

“A strongly opinionated minority can dictate group choice,” the research team wrote in its report, published in Thursday’s edition of the journal Science. “But the presence of uninformed individuals spontaneously inhibits this process, returning control to the numerical majority.”

The behavior of golden shiners demonstrates “the role of uninformed individuals in achieving democratic consensus amid internal group conflict and informational constraints,” wrote the research team, which included experts in biology, physics and engineering from Princeton, the University of Sussex in England, and the Max Planck Institute for Physics of Complex Systems in Germany.

In a separate commentary in the same issue, two authors from the University of Washington at Seattle, Carl T. Bergstrom, an associate professor of biology, and Jevin D. West, a biology research associate, said they agreed the work by Mr. Couzin’s group showed that “uninformed agents can promote democratic outcomes in collective decision problems.”

“Jefferson’s passionate arguments on the importance of education for democratic society notwithstanding,” the commenters wrote, the Couzin team has “identified circumstances in which ignorance can promote democracy.”

Such ideas, however, appear to enjoy less support among university researchers with actual experience studying how humans resolve political disputes.

“The claims they make are not ones made by political scientists,” said Lynn Vavreck, an associate professor of political science at the University of California at Los Angeles. “In fact, they are opposite.”

Human civilization, said Larry J. Sabato, a professor of politics at the University of Virginia, “is better off when more of its members are well informed and think carefully about the choices facing the society.”

Both Ms. Vavreck and Mr. Sabato said they could accept some broad analogies, such as the idea that voters can sometimes change directions quickly like a school of fish, and that vocal minorities can sometimes influence weakly committed members of a majority.

But they challenged the notion that people with less intense preferences are necessarily less informed, and the implication of Mr. Couzin’s study that a minority group is necessarily on the “wrong” side of a debate–citing, for example, civil rights protesters in the 1960s.

Mr. Couzin said in an interview that the study of animal behaviors to model various forms of social interaction among humans, in such areas as economic trends and consumer preferences, was well established.

But he acknowledged that human political relationships are far more complicated than the behavior of fish. “To apply this idea to specific systems,” Mr. Couzin said, “one would have to make specific models.”

This entry was posted in Uncategorized. Bookmark the permalink.

  • Print
  • Comment
  • 22072036

    I was an alumni interviewer for Georgetown University for years. The number of candidates I rated very highly who did or did not get in was about even, which did not surprise me given the competitiveness and huge numbers of applicants. What did surprise me were the instances where I rated someone low (explaining my rationale in detail) but yet that person was still offered a spot. That was when I felt my imput was not given much weight.

  • Brian Abel Ragen

    While I wonder why the applicants I rate most highly don’t get in, something else bothers me about alumni interviewing. I have been doing it for decades now, and somewhere along the line the applicants all turned pro. They arrive with C.V’s longer than my own–and sometimes with supporting documents. Everyone has been in elite summer programs and no one has failed to head at least one organization in high school. They all write thank-you notes after the interview.

    If they weren’t mostly still sweet kids, all this would make me want to gag. All I want is for the applicant to take the interview seriously enough to wear grown up clothes and talk about what they’re interested in. A polite, “Very good to meet you” on the way out is better than a “I’ve been trained to be a suck-up” note in the mail.

    In other words, I think applicants should be amateurs, not over-coached professionals who sometimes have been groomed by hired coaches or parents doing the same thing with a how-to book. (For an example of the latter, see _Getting Into College with Julia Ross_; you can find the former with a Google search–but don’t expect to find the fees posted.)

  • sand6432

    In the Dallas area we have almost 80 Princeton alumni interviewing over 350 applicants. I have been doing interviews for nearly two decades, first in State College, PA (Penn State), and now here, I usually interview three or four applicants each year. Only one (a valedictorian at State College High School) was admitted, but I’m not frustrated or annoyed. I feel that I am providing a useful service both to the University and to the students applying, and I benefit by getting to know some really outstanding young people from the area who will likely be successful wherever they go to college and whatever career they pursue later. The value to the Admissions Office, I understand, is in providing “texture,” as one college admissions official quoted in your earlier article put it. And it can make a difference in close cases (Stanford’s admissions officer quotes 10% of the 4,000 cases). So I am happy to continue participating. But I do worry about the increase in the number of applications owing to the adoption of the Common Application; I fear that some students are applying now almost pro forma, and i have had some not even show up for interviews, which never happened in earlier times. I also offer a word of caution to alumni interviewing recruited athletes: do not buy anything for them (like coffee at a Starbucks), or you may well be in violation of NCAA recruiting guidelines.—Sandy Thatcher

  • abomasho

    Peace, mercy and blessings of God
     
    God bless the Almighty in the government of King Abdullah bin Abdul Aziz
     

     Jiddawi

  • tylerjohn

    If the majority is always right, this is good news.  Otherwise…

  • phikaw

    Whatever generalizations might be made from this — training fish to prefer one color over another — to human preferences and consumer behavior, exactly what qualifies that to be called “democratic” in the political sense or “consensus” in the sense of a deliberative process for arriving at a conclusion? No objection to the idea that human political behavior has “natural” origins or explanations, but more a question about the phrase “the role of uninformed individuals in achieving democratic consensus
    amid internal group conflict and informational constraints,” in the researchers report. It seems that the study is about the formation of preferences, and it’s not clear to me anyway what this implies for democracy per se. Maybe I’m just missing something?

  • richardtaborgreene

    our ability to see implications in fish color preference for democracy—amazes here, who needs ignorance and illogic when our brains factor analyze our environment for patterns however ridiculous

  • solidagojuncea

    If I applied a study of fish behavior to conclusions about human politics, the editor wouldn’t even send my manuscript to the reviewers.  But if you’re from Princeton and the Max Planck Institute, they will not only publish it, they will find like-minded commentators.  I guess ignorant editors will mindlessly follow the lead of anyone from prestigious institutions.  Ergo, editors are fish.

  • philosophile

    We’ve always known that people are sheep. Now we learn that they are fish too. Big deal.

  • k_steiner

    Unless the study is reported incorrectly, the fish/voters did not differ in how well-informed they were. Each of the populations had an accurate understanding according to its own experiences. Perhaps, though, they might be said to have varied in their level of conviction about their understanding. I’ll refrain from snarky comments about Ivy-League science.

  • vceross

    Well, then, I guess we’re in very good shape here in the U.S.

  • dank48

    Speaking as an editor, I can’t see why you’d malign fish.

  • geochaucer

    Might I point out that this was a study of FISH? I mean, if this kind of logic is acceptable, CERN  could have saved several billions of dollars by building a minnow tank rather than the super collider and just inferring the existence of the Higgs from how the fish swam.

  • renellin

    No matter what side of the political spectrum you are on, we have all seen special interest groups dominate the scene by getting continuous coverage with the media and repeating themselves over and over. The likely analogy of fish behavior in this case to be indicative of human group behavior aside, There are so many issues that hardly touch some of our lives, and yet it is all that seems to be talked about to the point where people find themselves behaving in ways they did not intend just because they’ve been swept along.

  • renellin

    I think this may be the first discussion I have entered where I think I may agree with every single poster. Perhaps the article is just that simple. I’ll take this opportunity to say that ‘studies’ are given so much credence depending on what the reporter is hoping to sell, that I would favor a restriction somehow entailing references–you shouldn’t be able to report on a study without citing references to the actual raw data, how it was collected, and how the researchers arrived at their conclusion. I hate those headlines.

  • navydad

    I love good jokes like this. What a great parody. The idea of generalizing from how schools of fish go after different colored food to how people make voting decisions is obviously ludicrous, but it is a fun parody of some of the more silly conclusions some researchers make.

  • mbelvadi

    Candidate for an Ignobel Prize next year?

  • ebennett64
  • philosophile

    Maybe some commenters know the chapter in Jonathan Swift’s “Gulliver’s Travels,” where Gulliver visits the Island of Lagado and is taken through the “Grand Academy of Lagado.”

    I call the type of research reported here “Lagado science.”

  • jwestsci

    I want to clarify
    that neither Carl Bergstrom nor I believe that one can directly
    extrapolate Couzin’s results to human systems.  Like fish, humans differ
    in their preferences; however, humans also
    lie, proselytize and manipulate processes of information integration in
    numerous ways. As we wrote in our perspective, this can easily lead to
    results opposite to those observed in the fish system: “One might
    expect groups with uninformed members to be
    particularly susceptible to tactical behavior by minority
    subpopulations. If that tactical behavior involved some sort of active
    proselytizing to accelerate conversion to the minority opinion, one
    would be right.”  Moreover, many human institutions for
    collective decision-making such as democratic voting involve
    centralized information exchange and control. Therefore, as we stress in
    our perspective and in other places, these human decision processes are better represented
    by social choice problems than by strategic distributed
    computation problems as modeled by Couzin et al. See e.g. 

    http://www.msnbc.msn.com/id/45688354/ns/technology_and_science-science/#.TuuVfkpJ_th

    http://www.sciencenews.org/view/generic/id/336953/title/Uncommitted_newbies_can_foil_forceful_few

    Sincerely,

    Jevin West and Carl Bergstrom

  • rpm13

    There is a huge literature in experimental social psychology on minority and majority influence. There is also a growing interest in principles of collective behavior across species. Both of these bodies of work suggest to me that extrapolation, although not direct, may not be as complex as you indicate. As well, social choice problems and strategic distributed computation problems may be different points on a continuum rather than categorically different problems. In human groups the kinds of purely human considerations you and other posters have mentioned often cancel out across groups, such that the typical or average group follows general laws of collective behavior. At the same time, these results present an interesting puzzle because the situation for the new fish pits the majority against a natural tendency as well as a previously influential minority.

  • YourLittleBrother

    Please cancel my subscription.

    What’s a shame is your blog can’t even be used to wrap fish.

  • philosophile

    I think the problem many of us have with this study and others like it is revealed in your statement, “Like fish, humans differ in their preferences.”

    Fish don’t “prefer” anything; they just behave. Human beings both prefer and behave. That fact undercuts research based on animal models–at least on those animals in which preference–i.e. choice–cannot be clearly demonstrated.

  • vceross

    How do we know what fish do, whether they prefer or behave?  I think the fact that we really haven’t much of a clue about other beings is what gets in the way of research.  Until we are able to communicate meaningfully with another, whatever it happens to be, there’s truly no way of knowing.

  • philosophile

    My goodness, surely neurobiology tells us something about the difference between fish and humans!

  • vceross

    philosophile:  what does neurobiology tell us?

  • icouzin

    I should like to broadly support the views expressed by West and Bergstrom. 

    Our paper, as emphasized by the title, is investigating democratic consensus decision-making in animal groups. This is very different to democratic societies and other complex human organizations. 

    However one issue to note, regarding their post, is that individuals with stronger preferences can strongly accelerate their capability of converting uninformed individuals to their ‘opinion’ – in this case, desired direction of travel. See Couzin et al (2005) Nature. 

    However, returning to the main issue – our work reveals that, in animal groups making movement decisions, uninformed individuals, *or those who are well informed but have no preference regarding where to go*, can play a major role – specifically inhibiting a minority of individuals who may be attempting to exert control over the group direction (as for example, for their own benefit). 

    The underlying mathematics of this process is shown to exhibit commonalities among a range of model-types where individuals both influence, and are influenced by each other. However the only reference to humans (outwith setting the scene in the introductory paragraphs) is that, in light of the apparent generic nature of the principles, we conclude with our “results suggest a principle that may extend to self-organizing decisions among human agents.” That is all – not that they *do*. We don’t make such claims. However we do provide testable predictions for how groups may come to consensus.

    One other thing – we are not suggesting that ignorance is good. Gaining information is typically a good thing for decision-making, and doing so in groups has the advantage that noisy information can be integrated. However if information is costly to acquire, or individuals have otherwise limited access to information, as it is for many animals, then uninformed/those who lack strong preferences will tend to promote democratic consensus. One can see this as both a good or bad thing. Again we make no value judgement regarding the quality of preferences – we are investigating the underlying dynamics. We are now exploring the evolutionary consequences.

    Sincerely,

    Iain Couzin

  • samirahills

    The post is very informative. It is a
    pleasure reading it. I have also bookmarked you for checking out new
    posts. 

     

    Properties
    in Indirapuram

     

  • lyraew

    Happy Holidays!

    Colorado College will be closed for winter break from noon on 12/22/11 thru 1/8/12.
    I will return your email the week of January 9th.

    Best wishes for a wonderful new year.
    ~Lyrae