Thinking Critically about Evolutionary Theory

K. P. Mohanan
Center for Integrative Studies
Indian Institute for Science Education and Research (IISER), Pune
April 2011

Note to the Reader

This write up has two parts. Part I focuses on the creationism-evolution debate, and seeks to
unpack evolution into its component propositions. It was originally meant as preparatory reading
material for a class debate. Part II focuses on the task of constructing an explicit theory (as
opposed to an operational definition) of species. Distributed to the class after the debate, it
supplemented the debate, to clarify and consolidate some of the issues that emerged in class. The
connection between the two parts can be outlined in terms of the following sequence of class
activities:
1. Part I is distributed to students, with the instruction that they think about the tasks in it,
first individually and then in their affinity groups, and come prepared for a class
discussion.
2. Class session 1: We set up a debate between an actual evolutionary biologist (Priya Iyer)
and a non-biologist who adopts the persona of a creationist (me), with a molecular
biologist (Alok Srivastava) presiding. Students are encouraged to interrogate each of us.
3. Class session 2: In this round of debate, Dr Iyer and the students are the authors of a
journal article supporting evolutionary theory. I quit the creationist persona and adopt that
of a skeptical journal reviewer. Students have to convince me that each proposition of the
evolutionary theory is correct.
4. In the course of the debate, the class discovers that the claims of evolutionary biology that
we have been debating on cannot be critically evaluated unless the notion "species" is
clearly defined. There are problems with the textbook definition (interbreeding with fertile
offspring) but we decide to accept it tentatively.
5. In the course of the alleged evidence to support species divergence, we discover that the
term "species" is used in such a way that two distinct species interbreed to produce a novel
species. The concept of “species” implicit in this use logically contradicts the concept
“species” in the definition.
6. Hence we search for a better definition, or a theory of species. Students don't have the
background to pursue this task. Part II is meant to help them in this enterprise, while also
summing up and consolidating the points that emerged in the class discussion.

Part I: The Creation vs. Evolution Debate

Creationism vs. Evolutionism
There are those who support the idea of a divine creation of the world, and those who support
biological evolution. Emotionally charged battles between these two groups have been going on
ever since Darwin published his Origin of the Species. This controversy is salient in school
education as well as in politics in many parts of the world. It is important, therefore, that serious
students of science understand this issue properly, and be able to respond to it critically.
One can gain a sense of this debate by looking at the Wikipedia entry at
[Link] and J. R. Minkel’s
Scientific American article, “Evolving Creationism in the Classroom.”
([Link]
Minkel reports the American Republican Vice-presidential nominee and Alaska Gov. Sarah
Palin as saying that the Judeo-Christian version of creationism (the idea in the Bible that God
created Earth and its life forms a few thousand years ago) should “get equal footing with
evolution” in science classes. "Teach both," the article reports her as saying in a 2006 televised
debate. "You know, don't be afraid of information."
Presenting both “evolutionism” and “creationism” as “information” would be simply providing
expositions of the two positions, where a textbook or teacher says, “Those who support
creationism believe that X, Y, …Z are true, but those who support evolutionism believe that X’,
Y’, …Z’ are true.” There is no harm in such expositions. After all, no teacher would find it
difficult to tell her students, “Once upon a time, humans used to believe that the earth was flat,
and there are still many of them around.” But if the teacher presents one of these positions as the
Truth (whichever it may be) she would be requiring the students to believe what she believes to
be true. That would be a form of indoctrination, something opposed to the very core of the value
system that drives scientific inquiry.
To take a closer-to-home example, it would be perfectly legitimate for a teacher to tell her
students that Ptolemy believed that the earth is the unmoving centre of the universe, while
Copernicus believed that the earth rotates on a tilted axis and revolves around the sun. If she is a
teacher who values the students’ critical understanding of the issues, she may also provide an
exposition of the evidence and argumentation in support of the two positions. What a teacher
cannot do is dictate what the students should believe. After getting students to engage with the
relevant evidence and argumentation, the teacher should leave the students to decide for
themselves what they should believe.
To decide whether they should believe the creationists or the evolutionists, students need:
A) the critical thinking capacity to decide for themselves what to believe, what to reject as
false, and what to keep on hold for further evaluation; and
B) critical understanding of both positions, including the relevant evidence and
argumentation that bears upon the choice between them.
What follows is an attempt to help you develop a critical understanding of (the different versions
of) creationism (including the version called intelligent design) and (the different versions of)
the theory of evolution, such that you can decide for yourselves what position to take. The
ultimate intention is to nudge you to pursue questions on biological evolution and engage in
inquiry along novel paths.

Thinking Critically about Creationism

The website [Link] describes creationists as “…men and women who share a
belief in God and a conviction that He created us and the world in which we live. Creation
science, in its most general sense, is an effort to apply the scientific method to discover how God
created the Heavens and the Earth.”

Questions for you

Q 1 What do you think they mean “the scientific method”?
Q 2 What does it mean to say “God created the Heavens and the Earth?”
Q 3 When they say “heavens”, are they referring to what you see when you look at the night
sky, or to a place where you go when you die, which presupposes the assumption of a soul
that continues to live after the body dies?

2
Q4 The formulation of the goal of creationism as “an effort to apply the scientific method to
discover how God created the Heavens and the Earth” carries the presupposition that God
did create the Heavens and the Earth. How scientific is it to accept that presupposition
without critical scrutiny?
Q5 Why do you think I have raised questions 1-3?

Now take a look at their discussion of the issues in the creationism vs. evolutionism debate at
[Link] parts of which are given below:
Evolution vs. Creation: The Great Debate
The Evolution vs. Creation debate is often referred to as the "Great Debate." It's the emotion-packed question
of "Origins" -- why, how, and where did everything come from? 20th century science has made the compelling
discovery that, at some point, the universe began. Both sides of the Great Debate now agree that the universe
has not existed eternally. However, this is where the agreement ends. As far as the "why" and "how" of the
"origin event," this is where the division and contention begin. There are two basic theories in this Great
Debate. The first is the historical default - the Creation Model of Origins. This theory maintains that the
intricate design permeating all things implies a Designer. The second theory is the more recent, atheistic
explanation - the Evolution Model of Origins. This theory postulates that the intricate design permeating all
things is a product of random chance and excessive time.

Evolution vs. Creation: The Contentions

Evolution vs. Creation is indeed the Great Debate of our scientific times. In any scientific debate, the theories
must be tested according to the evidence. We propose that the burden of evidence should be upon the
Evolutionists, since Creation has been the historic and inherent default throughout virtually all cultures and
religions until roughly the last 200 years. Of course, Evolutionists, who view themselves as the only
"scientists" in the debate, insist that the burden of evidence be upon the Creationists. Evolutionists reason, we
cannot see the Creator, we cannot hear the Creator, and we cannot touch, taste or smell the Creator. Therefore,
we are unable to test for the Creator with any form of scientific equipment developed thus far. Creationists
retort, we cannot see, hear, touch, taste, or smell the human mind. We cannot test for the human mind with any
form of scientific equipment developed thus far. When we run an electroencephalogram, we are measuring salt
flow and electrical activity within the human brain. We cannot so much as even locate the human mind. Yet we
watch as human carcasses run about, making order of disorder, conscious decisions according to subconscious
criteria. We see the design and complexity that result from the operation of the brain through the invisible
realm known as the mind. Thus, we know with certainty that the human mind exists. Therefore, it's absolutely
logical for Creationists to postulate the existence of a Creator based upon the same "evidence." The design we
see all around us came from one, grand concept, and such a concept can only come from a complex Mind.
Furthermore, the mathematical and physical laws inherent in all things (including, most dramatically, the Laws
of Thermodynamics and the Law of Cause and Effect) effectively validate this evidentiary claim.
Evolution vs. Creation: Origins
In the Evolution vs. Creation conflict, Evolutionists do quite well in terms of theoretical science, but fail to find
empirical evidence. Evolutionists theorize that the universe, with all that it contains (space, time, matter and
energy), exploded from nothing. This is contrary to the First Law of Thermodynamics. Where did space, time,
matter and energy come from in the first place? Thus, for Evolutionists, the ultimate question of Origins
remains unsolved. To complicate the Evolutionary position, this original explosion of everything from nothing
is unable to explain all of the complexity and fine-tuning in the universe, including cosmic "voids" and
"clumps", retrograde motion of the galaxies, etc. Despite numerous problems, this explosion from nothing has
been dubbed the "Big Bang" and is the accepted theory among the majority of Evolutionists. Evolution is a
very unique "science." Typically, scientists observe evidentiary data and then formulate their conclusions.
Evolutionists have formulated their conclusion, and now look for the missing data.

Let us take a careful look at the charge the theory of evolution is not supported by evidence.

3
Null Hypothesis
Consider the proposal that “… the burden of evidence should be upon the Evolutionists, since
Creation has been the historic and inherent default throughout virtually all cultures and religions
until roughly the last 200 years.” This in effect says that creationism is the null hypothesis. How
would you react to this position? [Also, do all cultures and religions until roughly 200 years
subscribe to creationism? e.g. What about Buddhism, Advaita, …]
To respond to the question about the null hypothesis, you need to understand what the term
means in scientific inquiry.
A null hypothesis is a proposition that you take to be true if there is no evidence against it. Take
the question, “Is there a correlation between the length of a person’s nose and her ability to do
research in mathematics?” I doubt if there has been any research that investigates this question,
and hence it is unlikely that there is any evidence that supports or refutes this hypothesis. In the
absence of evidence for or against the claim, what should you accept as true? Should you
conclude that there is a correlation or should you conclude that there is no correlation? We
would all agree that the conclusion should be that there is no correlation. That is the null
hypothesis, what we accept as true in the absence of evidence to the contrary.
Not all choices among hypotheses involve null hypotheses. Suppose there is a correlation
between two properties, say, the tendency to suffer from constant nausea and the tendency to eat
cloves. Is one of them the cause of the other? We should consider the following possibilities.
Eating cloves causes nausea.
Nausea causes the desire to eat cloves.
A certain gene causes both nausea and desire to eat cloves.
In the absence of evidence, we would not be able to make a choice. That is to say, none of the
above is the null hypothesis.
A similar situation is found in mathematics. Suppose a mathematician, say, Ramanujam,
proposes a mathematical conjecture, and no one has been above to prove it. No one has been
able to prove that the conjecture is false either. Should we then take Ramanujam’s conjecture as
the null hypothesis and conclude that, given that it has not been proved to be false, the
conjecture is true? No. As far as mathematical conjectures are concerned there are no null
hypotheses. If the conjecture has not been proved to be true or false, we say that it remains a
conjecture, to be investigated further.
In criminal law, the null hypothesis (called “presumption”) is that the accused is innocent.
(“Innocent until proven guilty.”) As a result, the “burden of proof” is on the prosecution. All that
the defense has to do is to pick holes in the prosecution’s argument. In civil law, in contrast,
both parties in a controversy must present arguments in support of their respective positions, and
the judge chooses the position supported by the strongest argument. This is called ‘probative
weight’ in law. (The criteria of truth are different for criminal law and civil law.)
What I am pointing to is the distinction between situations where the null hypothesis logic is
legitimate and situations where the null hypothesis logic is illegitimate. In doing so, I am
pointing to the need to listen to the creationist criticism, evaluate it critically, and decide for
yourselves how much of the criticism you should accept. I would recommend that you spend
some time thinking about it.
An Alternative Version of Creationism
Let us proceed. Consider the following proposition:

4
 Creationism A. The universe we live in was created by a Divine Being outside the universe
who set it in motion with the Big Bang singularity, and designed in such a
way that it would evolve without her (= the Divine Being’s) intervention to
be the way it is now, including the humans and other creatures on Earth.
The Divinity in the above theory is non-interventionist. Once the universe is set in motion with
the Big Bang, she does not intervene or control its evolution.

Questions for you:

Q 6 Is Creationism A in conflict with the scientific theories of the origin and evolution of
the universe, the origin and evolution of the solar system and the planets, the physical
history of the earth, the origin and evolution of life on earth, and the origin and
evolution of the human species? If your answer is yes, identify and articulate the
conflict. If your answer is no, modify the theory minimally such that it would result in a
conflict.
Q 7 Is Creationism A a scientific theory? If it is, say what makes it a scientific theory. If it is
not, say why it is not a scientific theory.
Q 8 Rev. Tom Honey is a vicar (a priest) in England. Watch his 18-minute TED talk on God
and the Tsunami at
[Link] Is his theory of
Divine Creation identical to Creationism A? If it is not, identify the differences. Are the
two theories in conflict? If they are, say what the contradiction is.
Q 9 Go to Khan Academy at [Link] and watch the 12 minute video
on Intelligent Design and Natural Selection (in the section on Biology). Is Khan’s
version of Intelligent Design the same as either Creationism A or the Creationism of
[Link]? If they are not the same, are they in conflict?

Thinking Critically about Evolutionary Theory

Read the Wikipedia article “Evidence of Common Descent” at
[Link] and carefully examine the evidence
and argumentation it outlines in support of shared ancestry in evolutionary theory. When you
read it, it would be useful to bear in mind that the existence of analogical phenomena (e.g., same
skeletal design of hands in humans and chimpanzees, same structure of six branches in
snowflakes, belief in rebirth in ancient Greece and in ancient India…) can be explained in terms
of the following strategies:
Shared ancestry and divergent evolution ( = analogy as homology)
Shared substrate at the micro-level and/or environmental pressures
leading to convergent evolution at the macro level (= analogy as homplasy)
The spreading of a given pattern from a given source to some
other entity or domain. ( = analogy as borrowing)
Pictorially, these three strategies can be represented as:
Homology Homoplasy Spreading
Ax B’ C’ B’ C’

Bx Cx Bx Cx Bx Cx
figure 1 figure 2 figure 3

5
Take the emergence of writing in different cultures in human history. Take the x in the above
figures as trait x. In figure 1, cultures B and C have a single ancestor, A, which had writing, and
the writing was preserved (with or without modification) in the descendents. In figure 2, the
ancestors of B and C are B’ and C’, and neither had writing: writing emerged independently in B
and C, either because of (a) environmental pressures, or (b) the properties of the substrate
(human brain-mind / human society), or a combination. In figure 3, ancestor B’ had writing, but
C’ didn’t. The two cultures came into contact, and culture C borrowed it from B’ (or B).
There are many examples of cultural analogies of this kind, analogous traits that appear in
diverse cultures. They include incest taboos, ways of cooking, territorialism, weapons, speech,
belief in demons, belief in ghosts, moral codes, institutional punishment for socially undesirable
actions, music, dance, painting, mathematics, medicine, trade, money, religious institutions,
family, tool use, … the list goes on. Anthropologists call them cultural universals. (See the
University of California TV production “CARTA: Human Cultural Universals at
[Link] (CARTA is Center for Academic Research
and Training in Anthropogeny)

Questions for you

Q 10 What is the evidence to show that species undergo evolution, that is, changes of traits
over successive generations?
Q 11 Suppose it turns out that there is evidence for the claim that species undergo evolution.
It could still be the case that while species evolve (= change their traits), those changes
do not result in divergence (a split, such that a single ancestor species gives rise to two
distinct species). What is the evidence to show that species undergo divergence?
Q 12 Suppose it turns out that there is evidence for divergence. For instance, we might be
able to show that humans, chimpanzees and orangutans descended from a single
ancestor species, that all insects descended from a single ancestor species, and that both
lions and tigers descended from a single ancestor species. However, that some species
had a single ancestor species does not necessarily show that all species had a single
ancestor species. What is the evidence to show that all multi-cellular and mono-cellular
species existing today (and in the past) came from a single ancestor species?
Q 13 We use each of the three strategies outlined above (in figures 1-3) for different subsets
of analogical phenomena. Given this situation, for any explanation in terms of shared
ancestry to be convincing, it should show that alternative explanations in terms of
shared substrate or borrowing are not feasible. Given that the analogy of the camera eye
of humans and octopuses had independent origins (homoplasy), why is it necessary to
assume that the hand plan of humans and of chimpanzees had a shared ancestor hand
plan? Why couldn’t it be that different mono-cellular species appeared on the earth
independently of one another from the shared substrate of the large molecules, and these
distinct cell types gave rise to different species groups of multi-cellular organisms?
Q 14 The Wikipedia article cites comparison of the genetic sequence of diverse organisms as
evidence for common ancestry. Do the analogies of gene sequencing necessarily
demonstrate common ancestry?
Q 15 The article also cites fossil evidence as evidence for common ancestry. It says, “The
comparative study of the anatomy of groups of animals reveals structural features that
are fundamentally similar or homologous, clearly demonstrating phylogenetic and
ancestral relationships with other organism, most especially when compared with fossils
of ancient extinct organisms.” Does “fundamentally similar” mean they are
homologous? Does fundamental similarity “clearly demonstrate phylogenetic and
ancestral relationships”?

6
Q 16 The article cites ‘Biogeography’ as further evidence for shared ancestry. Does it? (Now
you do the interrogation.)

When you address these questions, it might be useful to bear in mind the distinction between (a)
shared ancestry and divergent evolution as an explanation for the observed patterns, (b) random
mutation and selection as the mechanism for species divergence. The second part comes into
play only for those cases where we have already accepted the first part. The existence of natural
selection does not constitute an argument against homoplastic explanation.
After having gone through the first round of reading-and-interrogating, it would be a good idea
to read the attached article on Waddington (Johanan W. S. Slack “Conrad Hal Waddington: The
Last Renaissance Biologist?”), as well as the attached article on Octopus by Simon Morris
(Simon Conway Morris “Consider the Octopus.) And while you are at it, reflect on the following
questions as well.

Q 17 Why do you think I asked you to read Morris’s article, and the article on Waddington?
Q 18 Consider how novel species emerge in the course of evolution. The standard model is
that of diversification, one species splitting into two or more. But what about fusion,
the possibility of a new species coming into existence because of a combination of two
or more species? Is there any evidence to show that such a process cannot exist? If
there is no such evidence, and if it turns out that fusion exists, what would the
evolutionary tree look like, and what challenges would it bring to the theory of
evolution?

More on the CARTA talks on Cultural Universals: points to ponder

1. Donald Brown refers to homology (figure 1 above) as vertical transmission and spreading
(figure 3 above) as horizontal transmission. Does horizontal transmission exist in biological
evolution? Why is the term “transmission” inappropriate for homoplasy?
2. Consider the concepts of change, development, and evolution. Change is a difference in a
given entity between two states of affairs at two different times. Development and evolution
involve changes that are perceived to have direction or at least non-reversible changes. In
biology, “development” and “evolution” refer to non-reversible changes within an organism
and within a population across successive generations respectively. Linguists, like
biologists, use the term development to non-reversible change within an individual
(language development), but they use the term history when it comes to changes in a
population across successive generation (historical linguistics), reserving the term “language
evolution” to refer to the emergence of language in human history. Now consider the
concepts of vertical and horizontal cultural transmission in the CARTA talks. How does
this connect to biology and language?
3. In physics, change is studied under dynamics, investigated using the formal tool of
differential equations. Does this suggest that the study of dynamical systems may be of value
to researchers in biology, linguistics and anthropology? Should biologists, linguists and
anthropologists start learning differential equations?

7
 Part II: Towards a Theory of Species

Preliminaries: Evolutionary and Non-Evolutionary Theories

A traditional system of education expects you to understand what textbooks and classrooms
assert as ‘knowledge’, accept what they tell you without questioning, and learn to apply what
you have understood. The previous handout, Thinking Critically about Evolution (Round 1),
was meant to make you doubt and question assertions in textbooks and classrooms, look for
evidence and argumentation, and accept the assertions only if you are satisfied that they are
rationally justified. These are habits of thought that education should help you develop to a point
that they become part of your very way of thinking. This handout follows up with a more
specific aim: it hopes to help you expand the range of concepts and modes of explanation when
thinking about evolution and the systematic relationship across species.
If you are not familiar with the relevant evidence for the theory of evolution, it is perfectly
possible that you might imagine (and believe) that all the biological species found on this planet
today have always existed, that they had no ‘origin’. It is equally possible that you might
imagine that every species did have an origin — that they didn’t exist before a certain point in
time. Let us dub these positions the no-genesis hypothesis and the genesis hypothesis
respectively. The first chapter of the Old Testament (Genesis), and Darwin’s theory of evolution
(The Origin of the Species), both subscribe to the genesis hypothesis, but provide different
accounts of genesis.
Those who adopt the genesis hypothesis have additional options to choose from:
A. Cause: Did a supernatural agency cause the origin of life (create it), or did life emerge
from natural causes, for instance, the self-organization of molecules? The Old Testament
and its descendent called Intelligent Design subscribe to the position of supernatural
causation, while evolutionary theory subscribes to natural causation.
B. Simultaneity of specio-genesis: Did all the species on the planet emerge simultaneously, or
did different species appear at different times? The Old Testament subscribes to
simultaneous specio-genesis, while evolutionary theory subscribes to non-simultaneous
specio-genesis.
C. Divergence: If we adopt the non-simultaneous view, the next question is: Can a single
ancestor species diverge into two or more descendent species in the course of time?
Evolutionary theory holds that such divergent evolution of species is possible.
D. Convergence: Independently of our position on divergence, another question is: Can two or
more ancestor species converge into a single descendent species in the course of time? The
standard evolutionary tree carries the implicit assumption that such convergence species
does not occur.
E. Explanation for analogies across species: There are at least four possible sources for the
existence of analogies across species at the molecular and morphological levels:
(1) shared ancestry (a pattern in an ancestor is inherited by the descendents);
(2) horizontal transfer (i.e., borrowing: a pattern in one species spreads to another);
(3) independent emergence (the same pattern emerges independently in different
branches of the evolutionary tree, attributable either to environmental forces or to
the self-organizational predispositions of the molecular substrate).
Standard evolutionary theory employs (1) as its staple mode, occasionally allowing
independent emergence stemming from environmental forces, but making no attempt to
explain analogy through the properties of the substrate.

8
From a historical point of view, we find that human curiosity about the nature of reality, and
about the origin of the physical universe, of life, and of humanity began with mythological story
telling. The Old Testament and the creation stories of the other cultures are examples of using
mythology to make sense of the world. At a later stage, we proceeded to rationally coherent
speculations. The Rig Veda, the Upanishads, Plato, Heraclitus, and Sankara are examples of
such metaphysical speculation. When conclusions based on systematically gathered evidence
replace metaphysical speculation, it is then that science begins. Such evidence-based rationality
employs two related strategies: hypothesis testing, and explanation-based conclusions. Darwin’s
approach to evolution was that of hypothesis testing. Our discussions on evolution have been an
attempt at nudging you to adopt the strategy of explanation-based conclusions.
The theory of specio-genesis in the Old Testament, as we said earlier, involves a supernatural
creator. This is inadmissible in scientific inquiry. Suppose we eliminate the creator from this
theory; we can state the scientifically admissible part of that specio-genesis as follows:
Non-evolutionary Genesis:
All species came to exist simultaneously, independently of one another.
Replacing the supernatural cause with a natural one, we may represent this view as follows:
natural cause

S-1 S-2 S-3 S-4… S-n species

Figure 1: Non-Evolutionary Genesis
In contrast, evolutionary theory advances the claim: “Each new species arises by descending,
with modification, from an ancestral species.” (Darwin 1859) We may state this as a weak claim
or a strong claim as follows:
Evolutionary Genesis:
Weak claim: At least some diverse species on the planet today evolved from a common
ancestor. (Evolutionary Genesis Version 1)
Strong claim: All the diverse species that exist on the planet today evolved from a single
common ancestor. (Evolutionary Genesis Version 2)
We may represent these two positions diagrammatically as follows:
natural cause

S-1 S-2 S-3 S-4 ancestors

S-1a S-1b S-1c S-2a S-2b S-4a S-4b descendents

S-1ai S-1aii S-2bi S-2bii S-4ai S-4aii …

Figure 2: Evolutionary Genesis Version 1

natural cause

SPECIES-0 ancestors

S-1 S-2 S-3 S-4 descendents

S-1a S-1b S-1c S-2a S-2b S-4a S-4b

S-1ai S-1aii S-2bi S-2bii S-4ai S-4aii …

Figure 3: Evolutionary Genesis Version 2

9
This way of unpacking the idea of evolution sheds light on the different positions that are
possible to take, and the differences and similarities between them. We can derive the Old
Testament theory from fig. 1 by changing the natural cause to a supernatural one. With similar
modifications to fig. 2 and fig. 3, we can derive various versions of Intelligent Design theories.
We could also have a combination of supernatural and natural causes, the supernatural cause
generating the ancestor, and natural causes governing the rest of evolution.
Common to both versions 1 and 2 is the idea that evolutionary divergence of traits within a
species, in the course of time, can lead to evolutionary divergence of species themselves. Now,
current evolutionary theories permit evolutionary convergence of traits, but not that of species:
Divergence Convergence
of traits of species of traits of species
YES YES YES NO

Figure 4 Figure 5
An evolutionary theory that allows both divergence and convergence of species (in addition to
the already accepted divergence and convergence of traits) is a perfectly plausible conjecture
that needs to be carefully investigated, not dismissed without consideration.
A word of caution. It might be useful to distinguish the following kinds of species convergence:
i. Convergence of two species into a single species through interbreeding, resulting in the
reduction of diversity.
ii. Convergence of two species into a single species through interbreeding, resulting in the
formation of a new species and hence in an increase of diversity.
iii. Independent evolution of two different ancestor species generating the same species,
resulting in greater overall diversity (but less than what it might have been had the
emergence resulted in distinct species.)
The term "convergent evolution" in the literature refers to the convergence of traits, not to any of
the convergences in (i)-(iii). This is what comes under homoplasy. The existence of (i) is and
(ii) are acknowledged in the field, but, as mentioned above, this is logically inconsistent with
the phylogenetic trees that carry the implicit assumption that such convergences do not exist.
And furthermore, as we shall see, they are both inconsistent with the standard textbook
definition of species.
In the discussion above, we reviewed possible models for a theory of evolution. The review is a
way to enlarge the space within which we can construct alternative evolutionary theories without
taking the classical one for granted.

There are many evolutionary theories of specio-genesis that contrast with the non-evolutionary
one. One conjecture that they all share is that of divergent evolution of species. If we accept
“evolution” as “changes resulting from successive cycles of reproduction,” the claim that unifies
all evolutionary theories is: A single ancestor species diverges into more than one distinct
species in the course of successive cycles of reproduction. To critically evaluate this claim, we
need to answer an important question:
Question 1: In the claim that species divergence happens in evolution (and species
convergence does not happen), what does the term “species” mean?
This question calls for either a definition or a theory of species, probably both.

10
Text Book Definition of Species
I have not seen a single satisfactory definition of species. One textbook definition of species is ‘a
population of organisms capable of interbreeding to produce fertile offspring’. This definition
does not work for at least three reasons. First, it says nothing about species that engage in non-
sexual reproduction.
Second, it leads to a contradiction. There exist well-known instances of what are called “ring
species”, in which:
populations interbreed (and produce fertile offspring)
A, B √
B, C √
C, D √
D, A √
But A, C X
B, D X
If the relation, “X belongs to the same species as Y,” is transitive, then populations A, B, C, and
D must belong to the same species. Yet, by the definition of species, A and C cannot belong to
the same species, nor can B and D. Thus, we end up with a logical contradiction.
The third problem is the logical inconsistency of the definition. Though rare, we do come across
instances of fertile interspecies hybrids; one such is that of dogs and wolves producing fertile
offspring. A number of such cases are documented at
[Link]
“ If the parent species are extremely closely related, they may produce fertile offspring. These are the
edge cases in which the biological species concept can become too rigid. The biological species
concept states that animals belong to a separate species if they cannot interbreed. So, if animals
supposedly from different species interbreed and produce fertile offspring, then according to the
biological species concept they should be one species. However, we usually don't consider them the
same species because they differ in other features, such as geographic location, appearance, behavior,
and genetics.
Matings between these species produce viable, fertile offspring:
Canis species: Domestic dogs (Canis familaris), wolves (Canis lupus, Canis rufus) and coyotes (Canis
latrans) can interbreed and produce fertile offspring:
o Evidence for domestic dog DNA in wild Siberian wolves (Vila et al. 2003)
o Evidence for domestic dog DNA in wild coyotes (Adams et al. 2003b)
o Evidence for red wolf-coyote hybridization (Adams et al. 2003a); Northeastern coyotes
may be the product of hybridization between Canadian wolves and Western coyotes.
Cichlid species in Lake Victoria are extremely variable, displaying 500 color morphs. The species are
isolated through mate choice, which is determined by coloration: mates choose each other by color
pattern. However, in recent years, human activity has caused the water of Lake Victoria to become
cloudy. In these areas, the cichlids can't differentiate between species. In these cloudy areas, bright
color morphs have disappeared and the fish have become similar and dull in appearance through
hybridization.” (Seehausen et al. 1997)
These discussions indicate that biologists are using some criterion other than that of the textbook
definition to categorize populations into species. Given this other criterion, we can reformulate
the textbook definition as a biological law instead of as a definition:
Prohibition of Interspecies Interbreeding: If two populations belong to two distinct
species, they cannot interbreed to produce fertile offspring.
This prohibition is stated as an absolute law (without exception). Given the examples cited
above, this formulation cannot be right. If we state the law in probabilistic terms, however, it

11
appears to capture a true generalization. I would therefore suggest that we state the insight as
three separate but related interbreeding laws:
Interspecies Interbreeding Laws
Law 1 a: If two populations belong to distinct species, the probability of their mating is low.
Law 1 b: The probability of interspecies mating producing an offspring is low.
Law 1 c: The probability of fertile offspring resulting from interspecies mating is low.
As in the case of species divergence and species convergence, this discussion raises the
following questions:
Question 2: In the claims that:
a) the probability of interspecies mating is low,
b) the probability of interspecies mating producing an offspring is low, and
c) the probability of fertile offspring resulting from interspecies mating is low,
what does the term “species” mean?

Other Problems
There are other logical problems in the standard evolutionary theory that go beyond the well-
known flaws of the textbook definition of species. One of them arises from the application of the
following universal law in development and evolution:
Law 2: Like begets like.
We all know that rabbits cannot give birth to mice and vice versa, and that the seed of a mango
tree cannot grow into a cashew tree. Parents and offspring cannot belong to distinct species. This
is the phenomenon of “Like begets like,” which is expressed as law 2.
Notice that the law is about species, not about traits, or about varieties within a species. Thus, it
does not say that parents and offspring cannot have different traits or belong to different
varieties. Also, within an Evo-Devo theory, this law holds on reproduction, the -Devo part.
When it comes to the Evo- part of the theory, we have a different story:
Evolution takes place in the course of a chain of successive reproductions.
Evolution can result in an ancestor and descendent belonging to distinct species. (fig. 4)
If no single parent-offspring pair is allowed to be distinct in their species membership, how can
a chain of such links result in an ancestor-descendent pair with distinct species memberships?
What we see here is a different manifestation of the transitivity problem of the relation “x
belongs to the same species as y” that we encountered in the ring species.

species-j species-k

Evolution: ancestor descendent

Reproduction: …parent offspring-parent offspring-parent offspring-parent

species-j species-j species-j species-j

Figure 6
If no individual link in the chain allows a mismatch between the parent’s species and the
offspring’s species, how can a chain of such links give rise to a new species?
It is possible to eliminate the logical contradiction by constructing a theory of species that
conforms to the law, “Like begets like,” in -Devo but not in Evo-. If so, our third question can be
formulated as follows:

12
Question 3: In the claims that:
a) parents and offspring cannot belong to distinct species, but
b) ancestors and descendents need not belong the same species,
what does the term “species” mean?

Uniformity Principle
One assumption that has always been at the center of astronomy and physics is the Uniformity
Principle. The principle was extended to geology by James Hutton, and was subsequently
popularized by John Lyles; it was available to Darwin, and was something he appealed to. The
assumption that species diverge (fig. 4) needs to be tested now against this uniformity
principle.A general version of the principle that applies to both space and time can be
formulated as follows:
Uniformity Principle:
Unless there is evidence to the contrary, it is reasonable to conclude that a general law that
holds on the present time and current location (here and now) also holds for a distant time
and distant location (there and then, whether past or future.)
The principle works as follows. Suppose someone tells us that there existed humans with wings
a million years ago. We have not seen a single winged human; so, based on samples from the
present, we formulate a law: "Humans have no wings." As the null hypothesis, this would apply
to the past as well. Unless there is evidence to the contrary, we are justified in concluding that
there were no humans with wings in the past either.
Likewise, suppose someone asked you whether the gravitation of a particular planet in a far
away galaxy that we have not observed obeys the inverse cube law, and not the inverse square
law. We have no evidence to conclude either way, so we must go by the uniformity principle:
the law that applies to our location (here) also applies to the far away location (there).
By the same token, we must adopt the following null hypothesis for evolution:
Null hypothesis: Like begets like in evolution.
By this null hypothesis, ancestors and descendents cannot belong to distinct species.
The apparent logical contradiction between evolution and development in fig. 6 lends support to
the null hypothesis. In the absence of positive evidence to the contrary, we must assume that the
null hypothesis holds, in which case the divergence-of-species hypothesis would be false. To
establish the credibility of any version of evolutionary theory, in other words, we must show that
divergence of species can be observed in the present, and that hence the null hypothesis is false.
The kind of evidence that textbooks and popular literature commonly present for the existence
of evolution in the present happens to be evidence for the divergence of traits within a species,
not of species themselves. The story of the peppered moth, for instance, is often used to
demonstrate observable evolution in the present. The population of this moth before the
industrial revolution was predominantly light-coloured; but as industrial soot began to
accumulate, dark-coloured moths began to appear, and soon became the most common form.
But this does not point to the light-coloured and dark-coloured moths belonging to distinct
species; it is only evidence for adaptation, which results in the divergence of traits, and then in
divergent varieties within the same species. To give evidence for the divergence of species, we
must show that the divergence in the traits of the two varieties result in the two populations
bifurcating into two distinct species.
In the research literature on the subject, unlike in popular literature and textbooks,, there is
considerable experimental evidence for present day species divergence, summarized by Joseph
Boxhorn in “Observed Instances of Speciation.” ([Link]
[Link]) It is interesting that many of the examples involve hybridization of different

13
varieties or even different species resulting in the formation of a new species. Speciation through
interspecies hybridization can be represented as follows:
Species A Species B

parent-i parent-j

Species C
Figure 7
Interestingly, this is an instance of the convergence of species (fig. 5) that is currently not part of
the standard theory. Given this result, we must revise question 1 as follows:
Question 1’: In the claim that both species divergence and species convergence happen in
evolution, what does the term “species” mean?

A First Step towards a Theory of Species

I am not a biologist, so I will not attempt to construct a theory of species: I will leave that task to
the experts in the field. As a trans-disciplinary theorist, however, I would like to sketch the bare
outlines of the kind of theory we are looking for, and to offer a couple of suggestions.
First, there is an important difference between an operational definition that we can use to
mechanically decide on the grouping of an observed entity, and a theoretical definition of an
abstract concept as an ingredient of an explanation. The definition of a woman as an adult
human female is an operational definition: given any living or non-living observable entity, we
can make a decision on its being a woman by asking “Is it a human being?” “Is it a female?” and
“Is it an adult?” The definition of a female as a class of organisms with ovaries is also one that
lends itself to a mechanical operation to decide on its femaleness. In contrast, a theoretical
definition may not be directly connected to decisions on observables. The connection between
the observables and the definition is made through the mediation of a set of theoretical
propositions. The construct that is defined this way resides at nexus of the propositions.
Unlike the distinction between male and female, for instance, the category of feminine is a
theoretical concept. Likewise, unlike speed, for which we can provide an operational definition,
no operational definition is possible for gravity. If we define gravity as the force that holds
between any two bodies with mass, we still need to connect the definition to the observable
entities through a number of other definitions (e.g., force, acceleration, velocity, mass, etc.), and
the laws of gravitation and motion.
The type of definition we are looking for in the concept of species is a theoretical one.
Something like the following is a plausible candidate:
Species (Def): Species-hood is an abstract property of a population of biological organisms
that guides the reproduction and evolution of the population.
This definition does not tell us whether or not we should group wolves and dogs into a single
species or two distinct species, but it is adequate for the purposes of acting as a starting point for
a theory. Next we must formulate a set of general laws in such a way that the concept of species
appears at the nexus of these laws, and thereby acquires empirical substance. The following laws
might be a reasonable starting point, even though they will have to be revised significantly:
Law 3a: Resemblance: The aggregate of equivalences (analogies) in genes, proteins,
anatomy, physiology, ontogenesis, and behaviour is greater within
species than across species.
Law 3b: Mating: The affinity for mating is greater within a species than across

14
 species.
Law 3c: Like Begets Like: Parents and offspring cannot belong to distinct species. (= law 2)
Law 3d: Divergence: When two varieties of a species are prevented from mating by
external conditions, they evolve into distinct species over time.
Law 3e: Hybrid Offspring: Mating across species is unlikely to produce offspring.
Law 3f: Fertile Offspring: Offspring resulting from mating across species is unlikely to be
fertile.
[Sidebar: Notice that Laws 3a, b, e, f are probabilistic laws. At some stage in the fleshing out of
the theory of species, it should be possible to formulate these laws in quantitative terms. In
particular, this is necessary for the law of resemblance, which cannot be critically evaluated
without designing a way of numerically calculating the distance between populations in terms of
their aggregate equivalences. We would also need a threshold of distance beyond which two
populations would be treated as distinct species. These technical tasks are beyond the scope of
the conceptual exploration we are currently pursuing.]

What Kind of Categories are Species?

In logico-mathematical terms, a category is a set. A set is a collection of entities. A collection
consisting of an apple, a baby, the number six, and all the students in a university is a set; so is
the collection of all the even numbers. Thus, the members of a set may or may not have anything
in common. A category is a set of entities that have something in common. The set of even
numbers forms a category, but the set consisting of an apple, a baby, the number six, and all the
students in a university does not constitute a category. A category is a natural class, not an
arbitrary class.
In this sense, species is a category. But what formal concept of set must we use to model of the
concept of species? To answer this question, we need to briefly review the available options.
The classical notion of set is that of a crispy set with a clearly defined boundary, such that the
definition of the set can tell us whether or not a given candidate belongs to that set. Such a set,
by stipulating sufficient and necessary conditions in discrete terms, divides the universe into
two, as in fig. 8, separating entities that are members of the set from those that are not.

member non-member

Figure 8: Classical Concept of Set

The set of even numbers is such a crispy set.
Discreteness is an essential characteristic of a classical set. The set of red objects, for instance,
does not form a classical set, because redness is a gradient attribute, not a discrete attribute. The
formalism that captures such categories is that of the fuzzy set. There are two properties that
distinguish a fuzzy set from a crispy set. One is a center that expresses prototypical attributes,
and the other is the absence of boundaries. At the center of the set of red objects for instance, is
the prototype of red, receding to less protypical and marginal reds when moving away from the
center, as in fig. 9: (the reader has to imagine the red receding gradually, without breaks).

center

15
 fuzzy margin

Figure 9: Fuzzy set with center

There are other ways of combining the discrete-gradient parameter with the parameter of the
presence and absence of a center. For instance, we could have a fuzzy set without boundaries,
but also without a center. Imagine, for instance, the rainbow spectrum where orange is gradually
replaced by red on the one side and yellow on the other. Here, orange is a set and so is yellow,
and in between orange and yellow are entities that could belong to either — orangish yellow or
yellowish orange, as in fig. 10. Here, there is no center, and there are no boundaries. Likewise,
we could have a series of inner and outer circles with crispy boundaries, as in fig. 11. Such a set
has both boundaries and a center.

Figure 10: Fuzzy set without center Figure 11: Crispy set with center
Let us look at one last model of sets. Consider the following propositions:
Condition 1: The minimum age for admission to the program is 20.
Condition 2: A candidate will be admitted if (s)he has:
(a) scored at least 90% in the DPLE test; or
(b) completed a four-year undergraduate program in a North American or
European university.
Condition 1 is a necessary condition for admission. But it is not a sufficient condition.
Conditions (2a) and (2b) by themselves are neither necessary nor sufficient. It is only when they
combine that they jointly tell us what belongs to the set and what doesn’t.
Now consider the following definition of scientific inquiry:
Condition A: Only those forms of rational inquiry that justify their conclusions on the basis of
independently corroborable observational grounds constitute scientific inquiry.
Condition B: Academic inquiries that employ the following are forms of scientific inquiry:
1) quantitive observations
2) experimental observations or
3) instrumental observations
Condition A is a necessary condition. Conditions B1-3 are sufficient conditions. A particular
form of scientific inquiry may satisfies none, some, or all of these three conditions. Inquiry that
satisfies all of them is a prototypical form of scientific inquiry. To the extent that it satisfies less
conditions, it becomes less prototypical.

non-typical typical
prototypical
Figure 12

16
What Fig. 12 shows is a prototype model of sets.
Given such a range of formal concepts, the choice from among them will have important
consequences for a theory of species. If we adopt the classical set, for instance, a population
either belongs to a given species or doesn’t belong to it; there is nothing in between. In a fuzzy
set with a center, on the other hand, membership is gradient: the attribute of belonging to a
species can range from 0 to 1, instead of being either 0 or 1. This would mean that two varieties
of a species can be located at two opposite margins, and can therefore be intermediate between
belonging to the same species and belonging to two different species. This is precisely the
consequence we want in cases that fit with fig. 7: interbreeding between two different species
producing an offspring that then becomes the ancestor of a novel species.
Likewise, consider the problem of the ring species. Suppose we take aggregate distance along
the parameter of analogies as the criterion for species assignment. If the distance between two
populations goes beyond a threshold, then they belong to different species. If so, fig. 10 allows
for species assignments of the following kind:
colours belong to the distance between them
orange and yellow the same species below the threshold
orange and red the same species below the threshold
yellow and red distinct species above the threshold
This is exactly the result we want in the ring species. In the classical set theory, if A and B
belong to the same set along a given dimension, and B and C belong to the same set, then A and
C must belong to the same set. This is not necessary in a fuzzy-prototype set.
Finally, consider the logical contradiction we encountered in the application of the Like-begets-
like law to development (where it should apply) and evolution (where it shouldn’t apply).
Suppose “x and y belong to the same set” is determined in terms of aggregate analogy distance
in a fuzzy-prototype set. Given this as the model for species, we would expect the distance
between parent and offspring in any given parent-offspring pair to be very small, but we would
also expect the distance to keep increasing over successive cycles of reproduction. If so, it
follows that the law “Like begets like” would apply in development but not in evolution. The
logical contradiction arising from the classical set disappears in the fuzzy-prototype set.
In sum, replacing the classical set with a fuzzy-prototype set in modeling the idea of species
allows us to escape all the problems of the classical species theory discussed in the previous
sections.

Issues to think about

Adopting the fuzzy-prototype model requires us to rethink the species concept. Consider, for
instance, the formulation of our laws of mating, hybrid offspring, and fertile offspring, repeated
below for convenience:
Law 3b: Mating: The affinity for mating is greater within a species than across
species.
Law 3e: Hybrid Offspring: Mating across species is unlikely to produce offspring.
Law 3f: Fertile Offspring: Offspring resulting from mating across species is unlikely to be
fertile.
Let us compare this with the textbook definition of species:
If two populations belong to different species, they cannot produce fertile offspring.
The textbook definition, as we have seen, is inconsistent with the findings of biologists, and as
formulated above, the law is false. Furthermore, it is contradictory to the idea of experiments of
interbreeding to create new species to provide evidence for evolutionary theory. This falsity and

17
logical contradiction both disappear when the insight is formulated as three laws within the
fuzzy-prototype model. The insight that the crossbreeding experiments show can be expressed as
follows:
When mating happens between two populations that are almost not members of the same
species, their offspring is likely to become an ancestor of a novel species.
This result should not be surprising. Creativity is the result of the coming together of two
distinct strands that are sufficiently far apart but not in total opposition. One might also
speculate (as one of the students in the class did), that the threat of extinction coming from near
non-fertility leads to specio-genesis.
If this account is along the right track, we might be pushed to a somewhat radical conclusion. In
terms of the mathematics of graph theory, a tree is a rooted non-cyclic graph. The graph in fig.
(13) is a tree, but those in (14) and (15) are not:

Figure 13 Figure 14 Figure 15

If we allow convergent evolution of species, as in fig. 5 and fig. 7, we are forced to adopt fig.
15. And if we adopt fig. 15 and allow cyclic graphs, it naturally raises doubts about the single-
root requirement. What underlies fig. 2 is the idea expressed by fig. 14. Exploring the possibility
of accepting the idea in fig. 14 for evolution would revolutionize the way we think about the
phylogenetic tree.
If the issues discussed above have aroused your interest, you could pursue further some of the
related ideas in the following articles:
(1) Doolittle, Ford W. (1999) “Phylogenetic Classification and the Universal Tree.” Science 284,
2124 (1999). [Argues against the traditional tree in figure 13.]
(2) Pennisi, Elizabeth (2008) “Modernizing the Modern Synthesis.” Science 11 JULY 2008 Vol
321. [Reports on the Alternberg conference on Evolution that questions many of the
assumptions of the mainstream evolutionary theory.]
(3) Fuhrmann, G. Y (1991) “Note on the Integration of Prototype Theory and Fuzzy-Set
Theory.” Synthese 86: 1-27, 1991. [Discusses precisely what the title says it does.]
(4) Wake, David, Varvalee H Wake and Chelsea D Specht (2011) “Homoplasy: From Detecting
Pattern to Determining Process and Mechanism of Evolution.” Science 331, 1032. [Has an
excellent discussion of different categories of homoplasy.]
(5) Slack, Jonathan M. (2002) “Conrad Hal Waddington: the last Renaissance biologist?”
Nature Reviews Vol 3 November 2001 889. [Has an excellent discussion of Waddington’s
revolutionary ideas, including that of the epigenetic landscape.]
(6) Morris, Simon Convey (2011) “Consider the Octopus.” Embo reports Vol 12 / No3 182.
[Discusses convergent evolution and homoplasy.]

18