SciencePediaReconstructing the sprawling, four-billion-year history of life is one of science's grandest challenges. Without a time machine, how can we untangle the relationships between the millions of species on Earth? The answer lies in deciphering the clues embedded in the organisms themselves—in their anatomy, genetics, and development. This article delves into the foundational concept used to read this history: characters and character states. It addresses the central problem of how biologists translate observable traits, from petal color to DNA sequences, into a robust hypothesis of evolutionary relationships. You will learn the 'alphabet and grammar' of evolution, exploring the principles that allow us to build and evaluate family trees of life. The first chapter, Principles and Mechanisms, will unpack the core concepts, from defining characters and states to the powerful logic of the parsimony principle. Following this, the chapter on Applications and Interdisciplinary Connections will demonstrate how this framework is used to reveal the tempo of evolution, test macroevolutionary hypotheses, and forge powerful links with fields like developmental biology.
Imagine trying to reconstruct the complete family tree of every person on Earth, but with a catch: you have no birth certificates, no historical records, only the people themselves. How would you begin? You might start by looking for shared features. Some features, like having a head, are useless—everyone has one. Others, like having a unique mole on your left ear, are also not very helpful for grouping people. But what about a specific, heritable trait, like a distinctive nose shape or a rare eye color that runs in a particular family? Suddenly, you have a clue. You have a thread to pull that might just unravel a branch of the great human family tree.
This is precisely the challenge and the adventure faced by evolutionary biologists. The living world is a vast library of stories, but the books are written in a language we are only just beginning to decipher. To read these stories—to reconstruct the history of life—we need to understand the language's alphabet, its grammar, and its narrative structure. This language is built from the traits of organisms, which we call characters and character states.
Let's start with the basics, using a simple, beautiful example: the color of a flower. If a botanist is studying a group of related plants, she might notice that some have white petals, some have yellow, and others have red. In the language of systematics, the general feature she is interested in—the "petal color"—is the character. The specific variations she observes—"white," "yellow," and "red"—are the character states.
Think of the character as the question, "What color are the petals?" and the character states as the possible answers. This simple idea is tremendously powerful. A character can be anything from the number of legs on an insect to the presence or absence of a backbone, or even the specific sequence of molecules in a gene. For a feature to be a useful character for building evolutionary trees, it must be heritable—passed down from generation to generation—and we must be confident that we are comparing "apples to apples" across different species. This concept of comparing the same feature, inherited from a common ancestor, is called homology.
More formally, we can think of a character as a function, , that takes a species (or taxon, ) from our study group and assigns it a specific state, , from a set of possible states, . The states must be mutually exclusive; a flower can't be both entirely red and entirely yellow at the same time. By carefully defining these characters and states, we create the fundamental alphabet we need to spell out evolutionary history.
Once we have our alphabet, we need to understand the rules of how one letter can change into another—the grammar of evolution. Are all changes equally likely? Consider our flower color character. Is it more likely for a red flower to evolve into a yellow one than a blue one? Without more information, it's hard to say. The biochemical pathways for pigments can be complex, and a single mutation might cause a drastic shift. In such cases, where we have no reason to assume a particular pathway of change, we treat the character states as unordered. A change from "red" to "yellow" is considered one evolutionary step, just as a change from "red" to "blue" is. Any state can, in principle, transform into any other state in a single step.
But now consider a different character: the number of leaflets on a compound leaf. The states might be . Is it plausible for a plant lineage that has always had single-leaflet leaves to suddenly evolve five-leaflet leaves in one jump? It’s possible, but perhaps less likely than a sequence of smaller changes: from one to two, then two to three, and so on. When we have a biological reason to believe that changes happen in a sequential fashion, we treat the character as ordered (or additive).
The justification for this choice isn't arbitrary; it's rooted in the very mechanisms of life, particularly developmental biology. Imagine a fish species that can have scales with zero, one, two, or three tiny bony ridges (lamellae). By studying how these fish grow, biologists might discover that the ridges form sequentially: an embryo first develops one ridge, which then bifurcates to form two, and one of those bifurcates again to form three. It's developmentally impossible to create three ridges without first creating one and then two. This is powerful evidence! It tells us that an evolutionary change from state (no ridges) to state (three ridges) isn't a single event, but a journey through states and . For such an ordered character, we assign a cost to the change that reflects this journey. The cost of going from state to state is simply . Changing from to costs evolutionary "steps," while changing from to costs only step. By observing how organisms are built, we learn the grammatical rules for how they can evolve.
With our alphabet and grammar in hand, we can finally start reading the stories of relationship. Our goal is to find monophyletic groups, or clades—groups that contain a common ancestor and all of its descendants. Think of your immediate family: you, your siblings, and your parents form a monophyletic group.
The key insight of modern systematics, a field known as cladistics, is that not all shared traits are equally useful for this task. We need to distinguish between ancestral and derived traits. A plesiomorphy is an ancestral character state, inherited from a distant ancestor. An apomorphy is a derived, or "new," character state that evolved within the group we are studying.
To use our family analogy again, having five fingers is a plesiomorphy for humans. We inherited it from our deep primate ancestors and beyond. You can't define your immediate family based on having five fingers, because your cousins have five fingers too, as do monkeys. A shared ancestral trait is a symplesiomorphy, and it tells us about ancient history, not recent relationships.
Now, imagine that a unique mutation arises in your great-grandfather that gives him and all his descendants—and only his descendants—strikingly blue eyes. This new, derived trait is an apomorphy. Because it is shared by a group of his descendants, it is a synapomorphy (a shared, derived trait). This is the gold standard of evidence! It is a true family resemblance that unites that specific branch of the family tree. Finally, if you alone develop a new trait, like the ability to wiggle your ears in a very specific way, that is an autapomorphy—a unique derived trait. It makes you special, but it doesn't group you with anyone else.
So, the central logic is this: to identify a true evolutionary group (a clade), we must look for the defining synapomorphies that unite it and set it apart from other groups.
This sounds straightforward, but nature is often a messy historian. Different characters can tell conflicting stories. One trait might suggest that species A and B are close relatives, while another trait suggests A is closer to C. How do we resolve these conflicts and find the single tree that best represents the true evolutionary history?
One of the most powerful tools we have is the Principle of Maximum Parsimony. It is a form of Ockham's Razor, the famous idea that, all else being equal, the simplest explanation is usually the best one. In phylogenetics, the simplest explanation is the one that requires the fewest evolutionary events, or the minimum total number of character state changes.
To apply this, we need a way to determine which character states are ancestral and which are derived. We do this by including an outgroup in our analysis—a species or group of species that we know is related to our study group (the "ingroup") but branched off earlier in history. The character states seen in the outgroup are assumed to be the ancestral condition for the ingroup. For example, if we are studying mammals, we might use a lizard as an outgroup. Lizards lay eggs, so we can infer that egg-laying (oviparity) is the ancestral state, and live birth (viviparity), which evolved within mammals, is the derived state.
Once we have our characters polarized, we can map them onto any possible tree and count the changes. Imagine we are comparing a platypus, a kangaroo, and a human, with a lizard as the outgroup. All three mammals have lactation (they produce milk), while the lizard does not. The most parsimonious scenario is that lactation evolved just once, in the common ancestor of all three mammals. The alternative—that it evolved independently in all three lineages—would require three changes, which is far less parsimonious. We can apply this logic to dozens or thousands of characters, and the tree that has the lowest total "parsimony score" (the fewest total changes) is our best hypothesis for the true evolutionary history.
The parsimony principle guides us toward the simplest story, but evolution doesn't always follow the simplest path. Sometimes, a character state that looks like a synapomorphy is actually a trick of history. This phenomenon, called homoplasy, can arise in two main ways: convergent evolution (where two unrelated lineages independently evolve the same trait) and evolutionary reversal.
An evolutionary reversal is a fascinating twist where a lineage reverts from a derived state back to an ancestral one. A classic example comes from creatures that adapt to life in total darkness. Imagine a group of crustaceans. Their ancestor, living in the open ocean, was unpigmented (ancestral state 0). A branch of this family moved into shallower, sunlit waters and evolved pigmentation (derived state 1). This pigmentation is a synapomorphy for the sun-loving members of the group. But then, one of these pigmented species colonized a deep, dark cave system. Over generations, with no need for sun protection, it lost its pigmentation, reverting to the unpigmented state (a change from 1 back to 0). For this cave-dwelling species, its lack of pigment is not a shared ancestral trait with the deep-ocean ancestor; it is a secondary loss, an evolutionary reversal. These reversals remind us that evolution is not a straight march of "progress" toward complexity; it is a pragmatic process of adaptation to local conditions, and traits can be gained just as easily as they are lost.
This brings us to a final, crucial point. While the principles of characters, states, and parsimony are mathematically and logically rigorous, applying them to the real world is both a science and an art. The very first step—defining characters and states—can involve difficult judgments.
Suppose we are studying a group of animals and measure the length of a particular bone. The measurements are continuous: cm, cm, cm, cm. To use this in a parsimony analysis, we must convert this continuous data into discrete character states. This is called discretization. But where do we draw the lines?
We could set a single threshold at cm, creating a binary character: "short" (state 0) vs. "long" (state 1). Or we could use two thresholds, say at cm and cm, to create three ordered states: "short" (0), "medium" (1), and "long" (2). Or we could even treat the two thresholds as defining two separate, independent binary characters.
As it turns out, these different "coding" choices can have a real impact on the outcome of the analysis. Choosing an ordered, three-state character might give more weight to the large gap between the small-boned and large-boned animals, potentially favoring one evolutionary tree over another. Creating two separate characters might effectively double the weight of this single feature, potentially biasing the result. There is no single, universally "correct" way to do this. The choice depends on the biologist's deep knowledge of the organism, its development, and its functional anatomy.
This is not a weakness of the method; it is a reflection of the beautiful complexity of the biological world. Reconstructing the tree of life is not a mechanical process. It is a profound act of scientific interpretation, blending rigorous logic with deep biological intuition. It is a conversation between the data and the biologist, a slow, careful, and endlessly fascinating deciphering of life's epic story.
We have spent some time understanding the machinery of character states—these little tags, these labels like "has feathers" or "lacks a backbone," that we attach to the creatures we study. It might seem like a simple game of sorting. But the real magic, the real beauty, begins when we use this seemingly simple tool to ask profound questions about the four-billion-year history of life on Earth. It’s like learning the alphabet; the goal isn’t just to know the letters, but to read the epic poems written with them. In this chapter, we’re going to read some of those poems.
The first great lesson of phylogenetics is that not all similarities are created equal. For centuries, we grouped organisms by overall appearance, which is why whales, with their streamlined bodies and fins, were long thought of as fish. The modern science of cladistics teaches us to be more discerning. It tells us that the key to uncovering evolutionary relationships lies not in any shared feature, but in a specific kind of shared feature: the shared evolutionary novelty, or what biologists call a synapomorphy.
Imagine you are a botanist who has discovered a group of new plants. You notice that two of them, and only those two, share a unique feature not seen in any of their relatives—say, a special kind of protective casing around their ovules. This shared invention is a powerful clue. It suggests that these two species descend from a recent common ancestor that first evolved this feature, and it unites them in a natural group, or clade.
This immediately illuminates one of the most common traps in evolutionary thinking: grouping organisms based on shared ancestral traits, or symplesiomorphies. Suppose a student observes that a strange deep-sea worm and a distantly related jellyfish both possess a simple, sac-like gut, and concludes they must be close relatives. This is a mistake. If the jellyfish is our outgroup—a reference point for the ancestral condition—then that simple gut is an ancient design. Using it to form a group is like arguing that humans and lizards form an exclusive club because they both have backbones. The trait is so old that it tells us nothing about the unique, more recent family history within the vast group of vertebrates. Cladistics is a science of history, and history is a story of change; we track relationships by following the trail of innovations, not by cataloging relics of a distant past.
Once we know to look for shared novelties, we face another problem. Nature is messy. One character—say, wing structure—might suggest one family tree, while another character—like tooth shape—might suggest a different one. How do we decide which history is correct?
Here, biology borrows a beautiful and powerful idea from physics and philosophy: Occam's Razor. The principle states that when faced with competing explanations, we should prefer the one that is simplest—the one that makes the fewest assumptions. In phylogenetics, this is called the principle of maximum parsimony. An "assumption" or a "complication" is an evolutionary event: a character state changing from ancestral to derived (a gain), or from derived back to ancestral (a loss). The most parsimonious tree is the one that tells the story of life with the fewest of these "plot twists."
Consider a biologist studying a group of newly discovered fungi. Based on their traits, two different hypotheses for their evolutionary tree are proposed. To decide between them, the scientist can map the character states of the fungi onto each tree and literally count the minimum number of evolutionary changes required to explain the observed patterns. If Hypothesis 1 requires 12 changes and Hypothesis 2 requires 15, the principle of parsimony directs us to favor the first tree. It is the more elegant explanation, accounting for all our observations with less evolutionary "effort". This doesn't guarantee it's the absolute truth—a more complex history is always possible—but it provides the most rational and testable starting point.
A phylogenetic tree is more than just a wiring diagram of relationships. When drawn to scale, it can become a profound chronicle of the evolutionary process itself. In a special kind of tree called a phylogram, the length of each branch is drawn proportional to the number of character state changes that are inferred to have occurred along that lineage. The tree becomes a "change-o-meter."
Let's imagine a tree of extremophilic archaea, microbes living in the most inhospitable places on Earth. The tree shows four species clustered together on short, bushy branches, indicating a relatively steady and slow accumulation of changes. But a fifth species, its closest relative, sits at the end of a very long, lonely branch. What does this tell us? It suggests that since this fifth species diverged from its ancestor, its lineage has experienced a dramatic burst of evolution. It has been living in the evolutionary fast lane, accumulating changes at a much higher rate than its cousins. This long branch is a signal in the data, a silent testament to a history of intense selection, rapid adaptation, or perhaps even a period where the organism's genetic proofreading machinery became less reliable. The humble character state, when summed up over millions of years, reveals the very tempo and rhythm of evolution.
This all sounds wonderfully objective. But before we can count changes or measure branches, we must first create our data. And this is where the biologist acts as both a meticulous scientist and a discerning artist. How do we translate the glorious, complex anatomy of an organism into a sterile grid of 1s and 0s?
Think of a botanist studying a strange appendage on a plant's stem. In one species it is fused to the stem; in another it is free but bears spines; in a third, it is free and has become a winding tendril. Is this a single character, "appendage type," with many different states? Or is it a composite of several independent characters—"fusion: yes/no," "margin: smooth/spiny," "tip: simple/tendrilled"? A good systematist will "atomize" the structure, breaking it down into its smallest, putatively independent evolutionary parts. The evolution of stem fusion is likely a different genetic and developmental story from the evolution of spines. By coding them as separate characters, we create more, and more precise, hypotheses to test on our tree. This careful definition of characters is one of the most crucial, and debated, steps in the entire process.
The challenges multiply when we work with fossils. They are our only direct window into the past, but they are often broken, distorted, and incomplete. What happens when our data matrix is filled with question marks? This can lead to strange and subtle artifacts. A well-known phenomenon is "stemward slippage," where a fossil with lots of missing data is artifactually pulled toward the base of the evolutionary tree. The parsimony algorithm, in its quest to minimize steps, will often resolve an unknown state ('?') as the ancestral state ('0'), because doing so requires no new evolutionary change. As a result, a fossil with many question marks can be interpreted by the algorithm as being artificially "primitive," placing it as an early offshoot even if its few known traits suggest it's more derived. This is a beautiful, cautionary tale about the dynamic interplay between our data's quality and our algorithm's logic. It's also why many scientists prefer character-based methods, which use the full matrix of states, over distance-based methods that first collapse all that rich information into a single number of overall difference, potentially obscuring such important details.
The power of character state analysis truly shines when it becomes a bridge to other fields of biology, weaving them together into a more complete picture of life.
A stunning example of this is the link between evolution and development (Evo-Devo). The way an organism grows from an embryo to an adult can be a powerful clue about its evolutionary past. Imagine we are trying to determine whether smooth or hairy leaves are the ancestral condition in a plant group. First, we apply the outgroup criterion: we look at closely related species, and they all have smooth leaves. This is strong evidence that "smooth" is ancestral. But we can go further. We can watch the hairy-leafed plants grow. We discover that as tiny seedlings, their very first leaves are perfectly smooth—the hairs only appear later as the plant matures. The congruence of these two lines of evidence—one from relatives (phylogeny), the other from an individual's life history (ontogeny)—is incredibly powerful. It's like finding the original draft of a poem scribbled out beneath the final, revised version.
Perhaps the most exciting frontier is using character states to test the grandest hypotheses of macroevolution. Can a single evolutionary invention change the fate of an entire lineage? Can growing a shell, or evolving wings, or becoming warm-blooded actually make a group more "successful"?
Modern computational methods, such as the Binary State Speciation and Extinction (BiSSE) model, allow us to tackle this question head-on. Imagine you have a complete, time-calibrated tree of life for a group. You can "paint" the branches according to a character state—for instance, red for lineages that have wings and blue for those that do not. The BiSSE model then analyzes this painted tree to estimate two separate sets of parameters: the rate of speciation () and extinction () for the winged lineages, and the corresponding rates for the wingless lineages. For the first time, we can ask with statistical rigor: Did the evolution of wings ignite an "evolutionary radiation" by increasing the birth rate of new species or by protecting them from the specter of extinction?
From a simple descriptive label, the character state is transformed into a variable in a mathematical model of life's diversification. We have journeyed from the simple act of observation to testing the engine of biodiversity itself. The character state is not just a label; it is the fundamental unit of information, the nucleotide in the language of evolutionary history, and with it, we are finally learning to read the great book of life.