第1章
The Wondrous Journey of Life's Evolution
Some Assembly Required has been hailed as a modern classic in evolutionary biology, with Bill Gates naming it one of his top five books of 2020. Neil Shubin, the paleontologist who discovered Tiktaalik roseae-the iconic "fishapod" with wrists and elbows that helped bridge our understanding of the water-to-land transition-takes readers on a captivating journey through life's 4-billion-year history. Unlike traditional accounts of evolution that focus solely on fossils, Shubin weaves together paleontology, embryology, genetics, and molecular biology to reveal how major evolutionary innovations actually emerged. The book's genius lies in showing that nature's greatest transformations rarely happen when we think they did-lungs existed before land animals, feathers before flight, and limbs before walking. Through Shubin's accessible storytelling, we discover how our bodies contain the remnants of ancient viruses, repurposed fish genes, and molecular battles that have shaped not just humans but all life on Earth.
第2章
The Arrow Between Fish and Land: Evolution's Hidden Pathways
In 1986, a simple projected slide showing a fish transforming into an early amphibian launched Neil Shubin's four-decade quest to understand life's great transitions. That small arrow between the two creatures represented a profound puzzle: how could a fish possibly evolve to walk on land? The transformation seemed to require simultaneous changes to virtually every body system-developing limbs instead of fins, lungs instead of gills, and new ways of feeding and reproducing. What good would legs be if you couldn't breathe air?
The answer lies in playwright Lillian Hellman's observation that "Nothing, of course, begins at the time you think it did." This insight captures one of evolution's most powerful concepts-the idea that biological innovations rarely emerge for their current purposes but are repurposed from preexisting structures.
Take lungs, for instance. When Napoleon invaded Egypt in 1798, he brought scientists including Etienne Geoffroy Saint-Hilaire, who discovered bichirs-fish with both gills and air sacs connected to their esophagus, allowing them to breathe air through holes in their skulls. Later explorers found other air-breathing fish in the Amazon and Australia, some with teeth matching 200-million-year-old fossils. These discoveries revealed that air-breathing fish had existed globally for eons.
Bashford Dean, an eccentric curator who studied fossil fish (and wore medieval armor on Manhattan streets), compared embryos with fossils and concluded that swim bladders and lungs develop similarly-both budding from the gut tube, differing mainly in position. Recent genetic research has confirmed this insight: the genes that build swim bladders in fish are identical to those that make lungs in both fish and humans.
Similarly, the evolution of birds from dinosaurs followed this pattern of repurposing. John Ostrom's controversial work connecting dinosaurs to birds was dramatically vindicated in 1997 when Chinese paleontologists revealed fossils of a dinosaur with unmistakable downy feathers. This discovery brought Ostrom to tears at a scientific conference, as his thirty years of work had finally been confirmed. Subsequent discoveries revealed that carnivorous dinosaurs had various feather coverings long before they could fly-from simple tubular shapes in primitive species to true feathers with central shafts in those most closely related to birds.
The evidence shows that feathers didn't evolve for flight but likely served other purposes in dinosaurs, such as display or insulation. As Darwin recognized, biological innovations never emerge during the great transitions they enable but are repurposed from preexisting structures. Revolution comes through evolution, following twisted paths with detours and dead ends, where inventions arise in one context and become repurposed in another.
第3章
Embryos Reveal Evolution's Secrets
Despite Linnaeus's unusual disdain for salamanders, calling them "foul and loathsome animals," these creatures became catalysts for understanding major biological transformations. In 1864, Auguste Dumeril received six unusual salamanders from Mexico City. These adults possessed feathery gills and aquatic features unlike any known salamander species. When Dumeril bred them, he was astonished to discover that some transformed into completely different terrestrial salamanders without gills. The difference was so dramatic it was as if "Dumeril had put chimpanzees in an enclosure one year and returned the next to find both gorillas and chimps happily cohabiting the cage."
This mysterious transformation raised profound questions about biological development that would help illuminate evolution's mechanisms. Early embryologists like Christian Pander and Karl Ernst von Baer discovered that all vertebrate embryos begin with the same three fundamental layers: the inner layer becoming digestive organs, the middle layer transforming into bones and muscles, and the outer layer developing into skin and nervous system.
A fortuitous labeling mishap led to another insight when von Baer couldn't distinguish between embryos of different species. He realized that while adult forms might appear vastly different, early embryonic stages showed remarkable similarities. These similarities extended to specific structures like gill arches that appear in all embryos with bony skulls, though they develop into different structures in different species.
Walter Garstang (1868-1949) developed a revolutionary perspective on development and evolution through his study of larvae, particularly salamanders. Most salamanders begin life as aquatic larvae with broad heads, flipper-shaped limbs, and feathery gills before transforming dramatically during metamorphosis. This water-to-land transition, which took millions of years in evolutionary history, occurs in just days during salamander development.
Dumeril's "magical" axolotls demonstrated that salamanders can follow different developmental pathways depending on their environment. Those in dry environments undergo metamorphosis to become terrestrial adults, while those in wet environments retain larval features into adulthood. From these observations, Garstang formulated a principle that small changes in developmental timing can profoundly impact evolution.
The ammocoete-a wormlike vertebrate-reveals an even more profound story about our evolutionary origins. The surprising answer to vertebrate origins came from sea squirts-stationary, rock-attached filter feeders that look nothing like vertebrates. Their free-swimming larvae, however, possess all three vertebrate hallmarks: a dorsal nerve cord, a connective tissue rod, and gill slits. Garstang proposed that the first vertebrates evolved when sea squirt development was arrested at the larval stage, preserving these crucial features into adulthood.
Even our human features may reflect developmental timing changes. Adolf Naef's striking photograph comparing neonate and adult chimpanzees revealed that juvenile chimps possess remarkably human-like features-large cranial vaults, erect heads, and small faces-that transform during development. Adult humans retain many juvenile chimp characteristics, suggesting human evolution involved developmental slowdown, with extended gestation and childhood preserving juvenile ancestral traits.
Beyond timing alterations, Julia Barlow Platt (1857-1935) discovered that specialized cells from the developing spinal cord migrate to form vertebrate-specific features throughout the body. This discovery revealed that complex evolutionary changes can arise from simple developmental shifts-altered timing and the origin of a new cell type together produced the vertebrate body plan.
第4章
The Genome's Hidden Conductors
Francis Crick and James Watson's discovery of DNA's double helix structure revealed how genetic information could both build bodies and pass to future generations. DNA's structure-strands of billions of bases (A, T, G, C) tightly coiled in each cell-creates an unbroken connection between all living things through time, making DNA a library of life's history within every creature.
Emile Zuckerkandl (1922-2013), who fled Nazi Europe with Einstein's help, pioneered comparing proteins across species. Using the era's limited technologies, he found that human and ape hemoglobins were more similar to each other than to those of frogs and fish. His work with Linus Pauling laid the foundation for molecular evolution.
Allan Wilson (1934-91), a mathematics prodigy from New Zealand, catalyzed these ideas into a scientific movement. His student Mary-Claire King made a breakthrough discovery that human and chimpanzee proteins were nearly identical-more similar than those between different species of mice or fruit flies. This paradox challenged scientists: if our proteins and genes were so similar to chimps', what created the significant anatomical differences between us?
When the Human Genome Project was announced in 2000, surprising patterns emerged: humans and flatworms both have roughly twenty thousand genes, while plants often have twice as many. Even more striking, genes that code for proteins compose less than 2 percent of the human genome, leaving 98 percent seemingly geneless. What does the rest of the genome actually do?
Francois Jacob and Jacques Monod, former French resistance fighters, discovered that genomes contain two components: genes that code for proteins, and shorter DNA sequences that act as switches controlling when genes activate. These switches tell genes when and where to be active. If genes are the ingredients in a recipe, these switches contain the instructions for when to add each ingredient.
Laura Lettice and her team discovered that mutations causing extra fingers (polydactyly) weren't in genes themselves but in switches nearly one million bases away. This control region is highly conserved across species with appendages, from humans to fish. DNA's complex folding brings distant switches close to their target genes, allowing tissue-specific control of development.
Two types of genomic changes drive evolutionary transformation. First, mutations in genes can produce new proteins. Second, and more critically for major evolutionary shifts, changes in genetic switches can alter when and where genes are active without changing the proteins themselves. When Berkeley researchers placed a snake's limb-control switch into mice, the mice failed to develop digits-revealing how snakes lost their limbs through switch modifications while maintaining normal gene function elsewhere.
This explains Mary-Claire King's earlier discovery that humans and chimps differ little in proteins but greatly in appearance-the difference lies in the switches controlling gene activity during development. Our DNA functions like an acrobatic maestro, constantly coiling and unfolding to orchestrate development through precisely timed gene activation patterns.
第5章
Beautiful Monsters Reveal Evolution's Patterns
Monsters have long fascinated those seeking to understand nature's workings. By the 1700s, Samuel Thomas von Sommerring recognized that "monsters" were disruptions of normal development. Darwin transformed this field by recognizing variation as evolution's essential fuel-without differences among individuals, natural selection couldn't operate.
Thomas Hunt Morgan's decision to work with fruit flies revolutionized biology. His team examined thousands of flies under microscopes, searching for mutations that could reveal genetic functions. Despite the lab's smell of rotting bananas, Morgan attracted brilliant scientists to his "Fly Room." His student Calvin Bridges discovered the famous Bithorax mutant-flies with four thoracic segments instead of two-and mapped the chromosomal region responsible.
Edward Lewis developed a powerful technique for studying Bithorax: removing large chromosome regions and observing effects, then adding fragments back sequentially to determine their functions. This revealed that Bithorax wasn't a single gene but a group arranged like pearls on a necklace, each controlling a different segment of the fly's body. Remarkably, the genes' positions on the chromosome mirrored the body's front-to-back organization.
By the 1980s, scientists isolated the Antennapedia gene responsible for flies growing legs from their heads and discovered many similar genes arranged sequentially on chromosomes, each active in different body segments. Remarkably, these genes weren't unique to flies-they appeared in earthworms, frogs, mice, and humans, revealing universal principles of body development.
These "Hox genes," arranged like beads on a string with positions matching body segments, opened a new world of understanding how bodies develop and evolve across the animal kingdom. When researchers deleted specific Hox genes in crustaceans, they created "beautiful experimental monsters" with transformed limbs. These weren't just laboratory curiosities-they mimicked natural diversity found in different crustacean species.
The genetic architecture that builds invertebrate bodies extends to vertebrates like us. In mammals, Hox genes provide genetic addresses to different body segments that form vertebrae and ribs. When researchers delete Hox11 genes in mice, segments that would normally form the sacrum develop as lumbar vertebrae instead. Similarly, by altering genetic addresses, scientists can create mice with ribs extending all the way to their tails.
These genes are redeployed across the body to build any segmented structure. In limbs, Hox genes provide genetic addresses to different regions-upper arm, forearm, and hand-just as they do along the body axis. Surprisingly, fish have these same "hand genes," but they make fin rays instead of digits. This revealed that the fin-to-limb transformation didn't require new genes but repurposed existing ones. Evolution's great revolutions often involve using ancient features in new ways rather than inventing entirely new genetic machinery.
第6章
Copy, Paste, Evolve: Duplication Drives Innovation
In the 17th and 18th centuries, exploring animal anatomy was as thrilling as geographic expeditions to distant lands. Felix Vicq d'Azyr (1748-94) was among the first to systematically compare anatomical structures across species, recognizing that forelimbs and hindlimbs were essentially copies of each other, following the same pattern, almost as if they were duplicated organs in a series.
In 1913, one of Morgan's students discovered a rare male fly with tiny eyes. In 1936, Calvin Bridges examined these small-eyed mutants and observed that chromosomes appear banded with alternating dark and light stripes. The small-eyed mutant had one extra-long chromosome where a segment of bands appeared to repeat-an early glimpse of gene copying as a biological mechanism.
Susumu Ohno, who translated protein structures into musical compositions, developed an ingenious method to study chromosomal differences between species. He photographed stained chromosomes from various mammals, cut out the images, and weighed them as a proxy for genetic material. His surprising discovery: despite widely varying chromosome numbers, the total weight of genetic material was virtually identical across mammal species. However, some salamander species contained 5-10 times more DNA than others, and up to 25 times more than humans. Ohno concluded this extra DNA was "junk"-duplicated genetic material with no apparent function.
The genome resembles a musical score where phrases repeat in different ways to create diverse compositions. Nature constantly duplicates and modifies-from DNA segments to entire genes and proteins. This pattern appears everywhere: hemoglobin exists in multiple forms for different life stages, each a copy modified for specific functions. Similarly, keratin proteins in nails, skin and hair are variations from a single duplicated ancestral gene. Color vision relies on three opsin proteins, each tuned to different wavelengths, all duplicated from a single ancestor.
Our enlarged brain-a signature human trait that nearly tripled in size from our australopithecene ancestors-evolved through gene duplication. Researchers identified NOTCH2NL, a gene active in human cortical tissue but absent in monkeys. This gene originated through duplication of a primitive NOTCH gene found in everything from flies to primates. Humans possess three NOTCH duplicates that sit end-to-end in our genome, creating an unstable region prone to further duplication or deletion. People with duplications develop larger brains while those with deletions have smaller ones, often showing symptoms of schizophrenia and autism.
Roy Britten discovered that 40-50% of calf and salmon genomes consisted of repeating sequences. Modern genome sequencing confirms his findings: more than two-thirds of our genome comprises repeated sequences. Barbara McClintock discovered regions where chromosomes would break apart, finding these break points moved around the genome-revealing that genes could "jump" from place to place. Despite initial skepticism from colleagues who thought her "crazy," McClintock's persistence paid off when jumping genes were later found in every species tested. We now know about 70% of our genome consists of these mobile elements.
第7章
The Genomic Battlefield Within Us
Seeds for Neil Shubin's work were planted during weekly visits with Ernst Mayr, one of evolutionary biology's founding fathers. During one meeting, Shubin mentioned Richard Goldschmidt's controversial "hopeful monsters" theory-the radical notion that major evolutionary changes could occur in a single generation through dramatic mutations. While most biologists believed evolution proceeded through small, gradual changes, Goldschmidt's view raised important questions about how complex innovations could emerge when they required hundreds of simultaneous genetic changes.
Vinny Lynch, a tattooed University of Chicago researcher whose childhood fascination with the Hudson River ecosystem nearly derailed his education, studies how major evolutionary innovations emerge. His focus: decidual stromal cells-beautiful cells with large red bodies surrounded by rich connective tissue that are key to mammalian pregnancy. Using advanced gene sequencing technology, Lynch discovered that transforming fibroblasts into decidual stromal cells involves activating hundreds of genes simultaneously.
This seemingly impossible coordination was achieved through jumping genes-the same type McClintock found in corn. A single mutation in a jumping gene created a progesterone-responsive switch that then spread throughout the genome as the gene duplicated and jumped to new locations. These jumping genes were eventually "neutered"-losing their ability to jump while retaining their useful regulatory functions-revealing how genomic conflicts between jumping genes and the rest of our DNA can drive evolutionary innovation.
At the placenta's boundary between mother and fetus sits syncytin, a molecular traffic cop controlling nutrient and waste exchange. When biochemists analyzed syncytin's DNA sequence, they made a startling discovery: it matched not other animal proteins but viruses, particularly HIV. Sometime in our evolutionary past, a virus invaded our ancestors' genome. Instead of hijacking our cells, the virus was "neutered"-its infectious components disabled while its cell-connecting properties were repurposed for placental development. Different mammals have different versions of syncytin derived from separate viral invasion events.
Even more remarkably, Jason Shepherd's lab discovered that the Arc protein, essential for memory formation, forms hollow spheres identical to viral structures when examined under an electron microscope. Genetic mapping showed all land animals possess the Arc gene while fish don't, suggesting an ancient viral infection occurred about 375 million years ago. This domesticated virus became essential for memory formation in land animals. The human genome contains roughly 8% viral remnants-over 100,000 "dead viruses"-some repurposed for functions like pregnancy and memory, others lying dormant. This genomic battlefield demonstrates how selfish genetic elements occasionally create new biological capabilities.
第8章
Evolution's Loaded Dice: When History Repeats
Stephen Jay Gould's popular lectures emphasized how contingent events like asteroid impacts dramatically altered life's trajectory. Gould argued that replaying "the tape of life" would yield entirely different outcomes. However, recent evidence suggests some evolutionary outcomes might be more predictable than Gould believed.
Sir Ray Lankester, a combative Victorian scientist, discovered that many species evolved through "degeneration"-losing traits rather than gaining them. Parasitic shrimp shed shells, eyes, and digestive organs; cave-dwelling creatures across different species independently lost the same features. Similarly, snakes, certain lizards, and caecilian amphibians all independently evolved long, limbless bodies. These biological "multiples"-parallel evolutionary solutions arising independently-reveal important patterns in how evolution works.
David Wake of UC Berkeley has dedicated his scientific life to salamanders-remarkable creatures that can regenerate limbs, hearts, and spinal cords. Many salamander species can shoot their tongues half their body length in less than two-thousandths of a second-faster than muscle contraction should allow. Wake discovered their tongues function like biological guns, repurposing gill bones as projectiles and abdominal muscles as retraction springs. Most remarkably, this complex adaptation evolved independently at least three times in unrelated salamander species.
When thousands of salamanders froze to death in Point Reyes National Seashore lakes during an unusually severe California freeze, Shubin and colleagues acquired them for study. Using a special staining technique that made the bodies clear and skeletons blue, they discovered that variation in salamander feet wasn't random. The same patterns of bone fusion and digit loss appeared across species from different continents. The sequence matched embryonic development-digits formed in the exact opposite order they were lost in evolution (last formed, first lost), and bones that fused were those that budded from each other during development.
This revealed a powerful insight: if you know how a salamander limb develops, you can predict how it will evolve. The construction process biases evolutionary pathways, making certain changes more likely than others. These developmental constraints explain why multiple species independently evolve the same features.
Caribbean islands serve as natural laboratories where lizards have independently adapted to similar environments. On each island, lizards specialized for different parts of trees evolved remarkably similar body forms-canopy dwellers are large with crests, trunk dwellers have short limbs and triangular heads, ground dwellers have long legs and large heads. DNA analysis revealed these aren't close relatives-each island's lizards evolved these forms independently from separate colonizing populations.
Similarly, marsupials in isolated Australia evolved forms paralleling placental mammals elsewhere-marsupial flying squirrels, moles, cats, and even extinct lions and saber-tooths. These natural experiments demonstrate that evolution isn't merely contingent but follows predictable pathways shaped by genetic recipes, developmental constraints, and environmental pressures. The dice are loaded, making certain evolutionary outcomes far more likely than others.
第9章
Life's Great Mergers: How Cooperation Builds Complexity
Sometimes the world isn't ready for new inventions. Leonardo da Vinci designed flying machines centuries before materials existed to build them. Similarly, fish with lungs and limbs evolved in water long before land was habitable. Timing is everything in innovation, whether in evolution, technology, or scientific discovery-as Lynn Margulis discovered when her revolutionary theory faced rejection from fifteen journals in the 1960s.
Margulis proposed that organelles like mitochondria and chloroplasts were originally free-living bacteria and blue-green algae that merged with other cells. Despite facing ridicule, Margulis persisted. When DNA sequencing technology advanced in the 1980s, her theory was confirmed: mitochondria were genetically related to oxygen-consuming bacteria, and chloroplasts to blue-green algae, not to their host cell nuclei. Every complex cell contains two families of life inside it. Margulis lived to see her theory confirmed before her death in 2011, maintaining throughout: "I don't consider my ideas controversial, I consider them right."
J. William Schopf's search for Earth's earliest life led him to Western Australia's 3-billion-year-old Apex Chert rocks. After years of painstaking work, microscopic filaments resembling blue-green algae colonies were found in these ancient rocks. Chemical analysis revealed these structures contained carbon signatures characteristic of living organisms. The filaments represented at least five different microbe types, including primitive photosynthesizers and methane metabolizers, showing life was already diverse 3.5 billion years ago.
The earliest multicellular fossils appear in 600-million-year-old rocks as simple impressions of ribbons, fronds, and disks. Bodies are individuals with beginnings and ends that can reproduce, whose parts work together to create a functioning whole with emergent properties beyond what individual cells could achieve alone. The human body's four trillion cells sacrifice individual autonomy to coordinate their growth, reproduction, and death for the greater functioning of the whole.
This coordination requires specialized molecular machinery-proteins like cadherins that both hold cells together and facilitate communication between them. These proteins are metabolically expensive to produce, which explains why bodies could only evolve after Margulis's merged cells provided the necessary energy through mitochondria.
The organization of bodies resembles Russian dolls: bodies contain organs composed of tissues made of cells with organelles, all containing genes. Over billions of years, different parts relinquished individuality to become components of greater wholes. While bodies and cells depend on highly controlled behaviors of their constituent parts, beneath that order lies conflict. Cancer exemplifies this conflict-cells breaking rules by dividing unchecked or refusing to die appropriately.
Revolutionary inventions often come from finding new combinations, like wheeled suitcases that emerged from combining two long-existing technologies. Nature works similarly through combination as a source of invention. In 1993, Spanish microbiologist Francisco Mojica discovered strange palindromic sequences in salt marsh bacteria that later proved to be a bacterial defense against viruses. This discovery led to CRISPR-Cas9, a revolutionary genome editing technology that can modify the DNA of plants, animals, and humans. The development of CRISPR-Cas followed evolution's ancient pattern of repurposing: a bacterial defense mechanism was adapted for human use in genome editing.
第10章
The Braided River of Life
On Christmas Day 2018, Neil Shubin escaped a summer blizzard to climb Mount Ritchie in Antarctica's Transantarctic Range. Standing atop the ridge, he surveyed an ice plateau larger than the continental United States. The mountaintops here poke through glaciers, exposing a vibrant layer cake of colors-reds, browns, and greens-that hold 400 million years of Earth's history. This polar region was once a giant tropical delta and later a place of intense volcanic activity.
From this vantage point, one might be tempted to see an orderly progression of evolutionary change in these rock layers. But evolution doesn't proceed in a linear sequence. Our minds tend to connect dots into narratives, like those cartoons showing a parade from monkey to human. The term "missing link" reinforces this misconception of evolution as a great chain.
While fish do appear before land creatures in the fossil record, the changes enabling terrestrial life arose earlier, in water-dwelling fish. Every major revolution in life's history follows this pattern-nothing begins when we think it does. DNA, ever twisting and at war with itself, provides fuel for evolution's changes. With 10% of our genome made of ancient viruses and only 2% our own genes, life's history flows like a braided river rather than a straight channel. Mother Nature repurposes ancient recipes to craft bewildering varieties of life.
That simple diagram that launched Shubin's career thirty years ago-a fish connected to an amphibian by an arrow-now seems naive. When you know how to look, you can see billions of years inside the organs, cells, and DNA of all living things, revealing our deep connections to all life on Earth.
