The Method of Multiple Working Hypotheses
by Thomas C. Chamberlin
A Commentary by Thomas M. Zinnen
From the title of this essay, I learned to ask not "What's
your hypothesis?" but rather ask "What are the testable
predictions of your competing hypotheses?" Anybody can
have one hypothesis. The challenge is in generating several hypotheses
that predict different outcomes that are testable in experiment.
Multiple is the key: not just one hypothesis, but two, or even
better, several competing hypotheses, are ideal. 'Working' emphasizes
the tentative nature of them all: they're designed to be tested,
The title is pregnant, the subtitle is elegant. "Parental
affection for a favorite theory" can cripple an inquiry.
Chamberlin, a geologist, wrote this essay in the 1890's. To my
ears it sounds late Victorian. But Mark Twain was writing at
this time, so pith and wit were possible.
Chamberlin lays out his thesis that the method of multiple working
hypotheses is useful in "investigation, instruction and citizenship"
or roughly equal to a common phrase describing the functions of
public universities: 'research, teaching and service'.
Chamberlin strikes a recurrent theme, a two-pronged fork in education:
Aristotle and Socrates; The encyclopedist and the inquisitor;
The answer and the question.
An irony: while science is supposedly about "the known"
it is to Chamberlin more interestingly viewed as the process of
making the unknown known. Science is not omniscience. And research
is far more interesting in fields where there's much left to be
Ever open-minded, Chamberlin describes three phases of intellectual
methods "thus far" and anticipates there will be more.
Now one hundred years later, one could ask: has the 20th century
produced further phases?
This also illustrates the difficulty of talking about "the
scientific method." Even a century ago there was no single
"method" of scientific inquiry. But Chamberlin asks
a key point: what are the strengths and weaknesses of various
methods of doing science?
Expectation of omniscience has been a heavy burden. It can lead
to "Explainaholism" a nasty disease. You'll see it
a lot on TV science shows and in museums. Rarely will someone
with Explainaholism really explain several competing ideas. No,
the earth goes around the sun and suffice it to show only evidence
that supports that correct conclusion.
Generating an explanation is not a problem. The problem is when
the emphasis is on "interpretation" rather than testable
prediction. This is the difference between scrutiny by reason,
and testing by experiment.
Is there a distinction between "tentative hypotheses"
and "tentative theories"? Chamberlin's states that
"It is not the slowness...but the thoroughness" of the
investigation that is the key quality of the scrutiny.
Chamberlin warns of coddling the toddling new explanation in a
gem of a phrase: "pressing of the theory to make it fit
the facts to make them fit the theory."
An agricultural allusion from a time when the country was still
primarily agricultural. To winnow chaff from the grain the mixture
is tossed in the air, and the lighter chaff is blown away by the
wind while the heavier grain returns to the winnow basket. To
sift dust from the grain both are passed over a screen that allows
the dust to fall through but retains the grain. (This differential
stepwise treatment is analogous to many methods of biochemical
purification.) This approach to epistemology assumes that truth
exists and humans merely discover it. Another approach is that
humans generate truth.
Ruling Theories Linger
Paragraphs 10 and 11
Rather than coddle an idea, he suggests another tack. Like a
civil engineer testing the earthquake-resistance of a model building,
you gotta "Shake It Til You Break It" in order to learn
Here the idea of tangible is key. Tangible comes from the Latin
'tangere' meaning "to touch." I take it to mean "testable
by experiment." One can argue ad nauseum how to answer "How
many angels can dance on the head of a pin?" because angels
are intangible. When scientists started to wrestle with natural
phenomena, with things empirical, then the power of combining
Experience with the Mental allowed the beginning of the ExperiMental
Approach to Testing Ideas.
Chamberlin introduces the theme of the freedom to inquire and
alludes to repression of the freedom. He does not state whether
the repression came from academics or from politicians or from
religious leaders. One can find examples in recent history for
A century ago Chamberlin offers a lament familiar to many students:
"The vitality of study quickly disappears when the object
sought is a mere collocation of dead unmeaning facts."
Chamberlin compares and contrasts the approach of the "ruling
theory" to that of "the working hypothesis." In
one phrase the difference between "support" and "test"
is key: are facts to be found to support the theory, or to test
the theory? This may be parallel to the distinction between trying
to "prove" a theory compared to trying to "disprove"
A Family of Hypotheses
Having only one hypothesis means it can rather easily slip into
being a "controlling idea." Encouraging several competing
ideas reduces this tendency. This avoids the 'parental preference'
that Chamberlin claims scientists would give to a sole hypothesis.
Chamberlin also points out that having multiple competing hypotheses
makes more sense in those complex cases where several factors
may work in concert to generate an effect or phenomena.
While Chamberlin gives a geological example, biologists can imagine
a multi-factor situation such as a biochemical reaction sensitive
to temperature and pH involving two reactants and a co-enzyme
and its cofactor.
"Whet the discriminative edge" of each hypothesis:
a phrase that may allude to "Occam's Razor": when faced
with two or more explanations that fit the available evidence,
one should prefer the simplest explanation. This is also called
the preference for parsimony.
Ironically, the words "science" and "scissors"
are both related to the Greek word "schizo" meaning
"to split." Chamberlin alludes to science as scissors:
two blades of inquiry are better than one for cutting into the
Ah, the power of simultaneous vision from several standpoints.
Stereoscopic vision for the mind for multidimensional probing.
Chamberlin makes a bold claim: experimental science can be both
analytical and synthetic, what we might call today both reductionist
Paragraphs 20 and 21
We speak one word at a time even while our minds imagine a dozen
dawning ideas simultaneously. How can one share these ideas without
a hopeless tangle of dangling phrases? Chamberlin alludes to
maturity, patience and practice as keys to communication. But
we can start by choosing our questions carefully. You can ask
"What's your hypothesis?" and encourage students to
think of only one possible explanation. Or you can ban the singular
and always ask "What are your hypotheses?" and thus
model and encourage students to consider many possible explanations.
In viewing the field of "science education" one could
get the impression that there is One Right Way to teach kids and
We Are Going to Find It. A century ago Chamberlin laments this
monorail mindset. If musicians need a wide repertoire, to Chamberlin
teachers need "a full array" of ways to approach different
learners with different states of mind.
Chamberlin contends that having but one
idea leads to preconceived notions and misperceived evidence.
He asserts that the mindset of multiple competing hypotheses
should be applied in social and civic life--to our personal choices
and our public policies.