Wednesday 23 December 2015
Tinkerer V. Artist
Why Would Evolution Produce Non-Essential Genes?
Evolution News & Views December 22, 2015 2:09 PM
Recent papers have tried to identify the subset of all genes in a genome that are essential for viability. In "The Indispensable Genome," Science Magazine considers this capability a turning point in biology:
Game-changing moments in functional genomics often reflect the development and application of powerful new reagents and methods to provide new phenotypic insight on a global scale. Three independent studies describe systematic, genome-scale approaches to defining human genes that are indispensable for viability, which collectively form the essential gene set. On pages 1092 and 1096 of this issue, Blomen et al. (1) and Wang et al. (2), respectively, report a consistent set of ~2000 genes that are indispensable for viability in human cells. Moreover, very similar results were obtained by Hart et al. (3). For the first time, we now have a firm handle on the core set of essential genes that are required for human cell division. This opens the door to studying the roles of essential genes, how gene essentiality depends on genetic and tissue contexts, and how essential genes evolve. [Emphasis added.]
This achievement follows on the heels of yeast studies where researchers found only 1/6 of its 6,000 genes to be essential. That's an astonishingly low fraction. What functions do essential genes perform?
The yeast essential genes encode proteins that drive basic cellular functions such as transcription, translation, DNA replication, cell division cycle control, and fundamental metabolism. Moreover, the yeast essential genes share several attributes that reflect their critical role in cellular life. For example, they are often conserved and evolutionarily constrained, are highly expressed, and encode abundant proteins that tend to form stable complexes and thus are rich in protein-protein interactions.
The landscape of essential genes in human cells can now be explored using the conceptual framework established in yeast.
All three studies on human cells found only 10 percent of the 20,000 genes in the human genome are essential. What is the other 90 percent doing?
Boone and Andrews, authors of the review, indicate that patterns in the human essential genome are similar to those in the yeast essential genome:
All three groups found that human essential genes are highly conserved, and much like yeast, they encode abundant proteins that engage in protein-protein interactions. The core set of human cell essential genes also tend not to be duplicated and appear to have increased evolutionary constraints, as they evolve slowly and are associated with fewer deleterious single-nucleotide polymorphisms. Although many essential genes are involved in fundamental biological processes including transcription, translation, and DNA replication, a substantial fraction remains functionally uncharacterized. Indeed, each analysis prioritized a wealth of uncharacterized genes whose essential roles are waiting to be explored.
That leaves about 18,000 "non-essential" genes to explore. One possibility is that they really are essential, too, when partnered with other genes. Yeast cells, for instance, can survive two non-lethal single mutations, but die when both occur together. This is called "synthetic lethality." Researchers have identified hundreds of thousands of these synthetic lethal interactions in yeast, they say. Initial studies in humans show the following pattern (notice the use of the phrase "functional information"):
Blomen et al. begin to address the extent of synthetic lethal interactions in human cells by screening a set of five nonessential genes with roles in secretion for synthetic lethal negative genetic interactions. They discovered an average of ∼20 synthetic lethal double-mutant interactions for a given nonessential gene, and these interactions tend to occur with functionally related genes. Even this relatively small genetic network suggests that the properties of the extensive genetic networks mapped for yeast are conserved and can now be mapped efficiently in human cells. The genetic network described by Blomen et al. ought to catalyze large-scale, collaborative efforts to map genetic interactions in human cells. Such an effort promises to enable functional annotation of the human genome, because genetic interaction profiles are rich in functional information and provide a quantitative measure of gene function.
These discoveries have the effect of raising the number of essential genes. If a cell can't survive a double hit on two interacting genes (a "synthetic lethal" condition), this indicates functionality even if each gene can take a hit on its own.
The studies of Blomen et al., Wang et al., and Hart et al. reveal the core essential gene set for human cells, setting the stage for the next wave of new genetic and chemical-genetic science that will take place directly in human cells. A future challenge will be to develop genetic tools, such as conditional alleles of essential genes, for exploring the terminal phenotypes and the various molecular mechanisms underlying the lethality associated with perturbation of different essential functions.
How often do double mutants occur in non-essential genes? If infrequent, synthetic lethals will be invisible to purifying selection. A non-essential gene can mutate and life will go on. Wang et al. say as much:
Essential genes should be under strong purifying selection and should thus show greater evolutionary constraint than that of nonessential genes. Consistent with this expectation, the essential genes found in our screens were more broadly retained across species, showed higher levels of conservation between closely related species, and contain fewer inactivating polymorphisms within the human species, as compared with their dispensable counterparts (Fig. 2, E to G). Essential genes also tend to have higher expression and encode proteins that engage in more protein-protein interactions.
Blomen et al. claim similar findings: essential genes show more conservation. They claim that "old" essential genes "emerged in premetazoans." But then, to their surprise, they found new essential genes incorporated into old existing functional genes:
Remarkably, the products of "new" essential genes are more often connected with old rather than other new essential gene products, suggesting that they largely function within ancient molecular machineries (fig. S9, B and C).
In PNAS, Rubin et al. looked for "The essential gene set of a photosynthetic organism." They identified "718 putative essential genes" for the photosynthetic lifestyle of a cyanobacterium.
There are certain limitations to the essentiality information determined here. Although we identified genes that are essential to the organism when individually mutated, they do not represent a minimal gene set. Essential processes for which there are redundant genes will not be discovered using an approach based on single mutants. In S. elongates, however, this complication is of lesser concern than in most other cyanobacteria because of its small genome size, which at a streamlined 2.7 Mbp, harbors little redundancy. In addition, the findings of essentiality reported here apply only to the specific laboratory conditions used and are likely to be different for a subset of genes under other growth conditions. Finally, because ncRNAs, regulatory regions, and other intergenic regions are much smaller, on average, than protein-coding genes, the essentiality calls for these regions are inherently of lower confidence than those made for protein-coding genes. Therefore, conclusions of essentiality for non-coding loci and to a lesser extent, protein-coding genes must be validated by targeted mutation before definitive statements can be made about their essentiality.
In short, the count of essential genes will vary by lab and researcher. This is definitely a work in progress, so measures of essential genes will need refinement with more study.
Contrasting Predictions
Let's integrate this information by contrasting the predictions of design and Darwinism about essentiality. The distinctions are not clear cut. On the one hand, Darwinians would like to see purifying selection acting to eliminate non-essential genes because, for large populations like yeast, they incur a metabolic cost. On the other hand, Darwinians have historically appealed to "junk DNA" and "vestigial organs" to explain things that do not appear essential, pointing to the weakness of purifying selection to eliminate mutations.
Design advocates, too, maintain competing expectations in tension. They would like to find functions for all gene activity to falsify the junk-DNA myth. But they would like to allow for functions beyond mere survival: functions that a designer with an artistic taste would create for beauty and pleasure.
So who's winning this debate on essentiality? It's too early to tell. Not enough is known yet. Based on experience with ENCODE and modENCODE, it seems likely that more functions will be found for everything in the genome (barring neutral or near-neutral mutations), but they will not always be essential for survival. There is a wealth of phenotypic evidence to support this: the beautiful spirals of a conch shell, elaborate patterns in fur and feathers, and other cases of elegant design that seem to go beyond the requirements for reproduction. Animals could satisfy Darwin's criteria by being all gray and just getting by till they have offspring, but life is incredibly vibrant with "useless" beauty. One suspects that genotypic evidence will follow suit.
Boone and Andrews point to "a wealth of uncharacterized genes whose essential roles are waiting to be explored." Who is better prepared to explain what the "substantial fraction" of genes that remain "functionally uncharacterized" do -- those who start with the assumption that "if it works, it's not happening by accident" or those who expect cobbled-together bits of junk?
Evolution News & Views December 22, 2015 2:09 PM
Recent papers have tried to identify the subset of all genes in a genome that are essential for viability. In "The Indispensable Genome," Science Magazine considers this capability a turning point in biology:
Game-changing moments in functional genomics often reflect the development and application of powerful new reagents and methods to provide new phenotypic insight on a global scale. Three independent studies describe systematic, genome-scale approaches to defining human genes that are indispensable for viability, which collectively form the essential gene set. On pages 1092 and 1096 of this issue, Blomen et al. (1) and Wang et al. (2), respectively, report a consistent set of ~2000 genes that are indispensable for viability in human cells. Moreover, very similar results were obtained by Hart et al. (3). For the first time, we now have a firm handle on the core set of essential genes that are required for human cell division. This opens the door to studying the roles of essential genes, how gene essentiality depends on genetic and tissue contexts, and how essential genes evolve. [Emphasis added.]
This achievement follows on the heels of yeast studies where researchers found only 1/6 of its 6,000 genes to be essential. That's an astonishingly low fraction. What functions do essential genes perform?
The yeast essential genes encode proteins that drive basic cellular functions such as transcription, translation, DNA replication, cell division cycle control, and fundamental metabolism. Moreover, the yeast essential genes share several attributes that reflect their critical role in cellular life. For example, they are often conserved and evolutionarily constrained, are highly expressed, and encode abundant proteins that tend to form stable complexes and thus are rich in protein-protein interactions.
The landscape of essential genes in human cells can now be explored using the conceptual framework established in yeast.
All three studies on human cells found only 10 percent of the 20,000 genes in the human genome are essential. What is the other 90 percent doing?
Boone and Andrews, authors of the review, indicate that patterns in the human essential genome are similar to those in the yeast essential genome:
All three groups found that human essential genes are highly conserved, and much like yeast, they encode abundant proteins that engage in protein-protein interactions. The core set of human cell essential genes also tend not to be duplicated and appear to have increased evolutionary constraints, as they evolve slowly and are associated with fewer deleterious single-nucleotide polymorphisms. Although many essential genes are involved in fundamental biological processes including transcription, translation, and DNA replication, a substantial fraction remains functionally uncharacterized. Indeed, each analysis prioritized a wealth of uncharacterized genes whose essential roles are waiting to be explored.
That leaves about 18,000 "non-essential" genes to explore. One possibility is that they really are essential, too, when partnered with other genes. Yeast cells, for instance, can survive two non-lethal single mutations, but die when both occur together. This is called "synthetic lethality." Researchers have identified hundreds of thousands of these synthetic lethal interactions in yeast, they say. Initial studies in humans show the following pattern (notice the use of the phrase "functional information"):
Blomen et al. begin to address the extent of synthetic lethal interactions in human cells by screening a set of five nonessential genes with roles in secretion for synthetic lethal negative genetic interactions. They discovered an average of ∼20 synthetic lethal double-mutant interactions for a given nonessential gene, and these interactions tend to occur with functionally related genes. Even this relatively small genetic network suggests that the properties of the extensive genetic networks mapped for yeast are conserved and can now be mapped efficiently in human cells. The genetic network described by Blomen et al. ought to catalyze large-scale, collaborative efforts to map genetic interactions in human cells. Such an effort promises to enable functional annotation of the human genome, because genetic interaction profiles are rich in functional information and provide a quantitative measure of gene function.
These discoveries have the effect of raising the number of essential genes. If a cell can't survive a double hit on two interacting genes (a "synthetic lethal" condition), this indicates functionality even if each gene can take a hit on its own.
The studies of Blomen et al., Wang et al., and Hart et al. reveal the core essential gene set for human cells, setting the stage for the next wave of new genetic and chemical-genetic science that will take place directly in human cells. A future challenge will be to develop genetic tools, such as conditional alleles of essential genes, for exploring the terminal phenotypes and the various molecular mechanisms underlying the lethality associated with perturbation of different essential functions.
How often do double mutants occur in non-essential genes? If infrequent, synthetic lethals will be invisible to purifying selection. A non-essential gene can mutate and life will go on. Wang et al. say as much:
Essential genes should be under strong purifying selection and should thus show greater evolutionary constraint than that of nonessential genes. Consistent with this expectation, the essential genes found in our screens were more broadly retained across species, showed higher levels of conservation between closely related species, and contain fewer inactivating polymorphisms within the human species, as compared with their dispensable counterparts (Fig. 2, E to G). Essential genes also tend to have higher expression and encode proteins that engage in more protein-protein interactions.
Blomen et al. claim similar findings: essential genes show more conservation. They claim that "old" essential genes "emerged in premetazoans." But then, to their surprise, they found new essential genes incorporated into old existing functional genes:
Remarkably, the products of "new" essential genes are more often connected with old rather than other new essential gene products, suggesting that they largely function within ancient molecular machineries (fig. S9, B and C).
In PNAS, Rubin et al. looked for "The essential gene set of a photosynthetic organism." They identified "718 putative essential genes" for the photosynthetic lifestyle of a cyanobacterium.
There are certain limitations to the essentiality information determined here. Although we identified genes that are essential to the organism when individually mutated, they do not represent a minimal gene set. Essential processes for which there are redundant genes will not be discovered using an approach based on single mutants. In S. elongates, however, this complication is of lesser concern than in most other cyanobacteria because of its small genome size, which at a streamlined 2.7 Mbp, harbors little redundancy. In addition, the findings of essentiality reported here apply only to the specific laboratory conditions used and are likely to be different for a subset of genes under other growth conditions. Finally, because ncRNAs, regulatory regions, and other intergenic regions are much smaller, on average, than protein-coding genes, the essentiality calls for these regions are inherently of lower confidence than those made for protein-coding genes. Therefore, conclusions of essentiality for non-coding loci and to a lesser extent, protein-coding genes must be validated by targeted mutation before definitive statements can be made about their essentiality.
In short, the count of essential genes will vary by lab and researcher. This is definitely a work in progress, so measures of essential genes will need refinement with more study.
Contrasting Predictions
Let's integrate this information by contrasting the predictions of design and Darwinism about essentiality. The distinctions are not clear cut. On the one hand, Darwinians would like to see purifying selection acting to eliminate non-essential genes because, for large populations like yeast, they incur a metabolic cost. On the other hand, Darwinians have historically appealed to "junk DNA" and "vestigial organs" to explain things that do not appear essential, pointing to the weakness of purifying selection to eliminate mutations.
Design advocates, too, maintain competing expectations in tension. They would like to find functions for all gene activity to falsify the junk-DNA myth. But they would like to allow for functions beyond mere survival: functions that a designer with an artistic taste would create for beauty and pleasure.
So who's winning this debate on essentiality? It's too early to tell. Not enough is known yet. Based on experience with ENCODE and modENCODE, it seems likely that more functions will be found for everything in the genome (barring neutral or near-neutral mutations), but they will not always be essential for survival. There is a wealth of phenotypic evidence to support this: the beautiful spirals of a conch shell, elaborate patterns in fur and feathers, and other cases of elegant design that seem to go beyond the requirements for reproduction. Animals could satisfy Darwin's criteria by being all gray and just getting by till they have offspring, but life is incredibly vibrant with "useless" beauty. One suspects that genotypic evidence will follow suit.
Boone and Andrews point to "a wealth of uncharacterized genes whose essential roles are waiting to be explored." Who is better prepared to explain what the "substantial fraction" of genes that remain "functionally uncharacterized" do -- those who start with the assumption that "if it works, it's not happening by accident" or those who expect cobbled-together bits of junk?
On our neighbours' minds.
Furry, Feathery, and Finny Animals Speak Their Minds
Denyse O'Leary December 22, 2015 3:27 AM
Much research on animal minds is rooted in Darwinian naturalist assumptions -- a long slow continuum of intelligence from somewhere just north of cytoplasm to humans. These assumptions may have set us back. First, just being a life form includes a drive to survive and an ability to adapt for that purpose, which we do not find in rocks. Many life forms can also communicate for those purposes. But, so far as we know, they lack consciousness or sentience, the ability to feel things. There is no slow ascent; there is a steep cliff.
At the other end of the spectrum are apes, who belong to the same order of life as ourselves. Discussions of their intelligence often assume that they are entering a "Stone Age" (such as the Lascaux cave artists lived in, 20 000 years ago). However, while apes tend to be more intelligent than most mammals, they are not becoming like humans. And smart birds give apes serious competition, when tested.
So what can we learn from other vertebrate life forms, forms that show intelligence but are not closely related to us, do not seem much like us, and are not apparently heading in our direction?
Can Animal Mind Be Explained by "Instinct"?
At one time, it was supposed that most animals were simply born with instincts about what to do. The term is not used much now because it mainly meant that we do not know the source of the animal's information. We are now learning many of these sources.
We have recently discovered, for example, that migrating birds can use the mineral magnetite, embedded above their beaks, to use Earth's magnetic fields for direction.
We learned in recent decades how young birds "know" that they should follow their mother: As Spark Notes explains, following the work of animal behaviorist Konrad Lorenz (1903-1989):
Johnson and Bolhuis identified two independent neural systems that control filial imprinting in precocial birds. Newly hatched chicks will follow almost anything that has eyes and moves. After the chick follows something, another part of the brain, analogous to the frontal cortex, recognizes and imprints on the individual being followed. These mechanisms are independent. There is an instinct for chicks to follow, and then they learn what they are following.
But the "follow her" system is not strictly genetic:
It might seem odd that being able to identify and follow a mother does not have a genetic mechanism. Yet with a neural rather than genetic mechanism, the chick gains flexibility that might help in survival. If a chick's mother dies, the chick can then be adopted by another family member or conspecific.
Yes indeed. Famously, the young bird may follow a psychology student, a stick, or a cat, with varying results. Bird rescuers use hand puppets of bird faces when caring for nestlings, to return them later to a natural setting. Clearly, not all we need to know about an organism is in its genes.
How then does the male weaverbird know how to build a nest? That's apparently not simply a genetic program either; the birds must learn some of the techniques by experience.
Genetics, neural networks, and experience all make animal learning much more complex and information-rich than the concept of "instinct" implied. But we are not yet in the realm of "intelligence." The migrating and nest-building birds access existing solutions to longstanding problems; they do not come up with new ones.
Both birds and mammals can learn to solve new problems presented to them. Let's look at some recent finds in mammals first, bearing in mind that we have only really begun to look at their intelligence seriously. It is early days yet, so some sketchiness is inevitable.
Mammals' Unexpected Intelligence
We find intelligence where we did not expect it. Pigs, for example (despite their reputation), are "socially complex as other intelligent mammals, including primates" (Natural History Magazine). That is surprising because pigs don't usually form close relationships with humans, as dogs do. And hog farming operations don't encourage intelligence.
We know more about the intelligent animal we are close to. In terms of communication, horses' surprisingly varied facial expressions are more similar to those of humans on one measurement than those of chimps are:
The Equine Facial Action Coding System (EquiFACS), as devised by the Sussex team in collaboration with researchers at the University of Portsmouth and Duquesne University, identified 17 "action units" (discrete facial movements) in horses. This compares with 27 in humans, 13 in chimps and 16 in dogs.
That might account for the human-horse bond (there seems no similar chimp-horse bond).
Horses', dogs', and cats' tail communications are also easy to read (they are intended to be). So just as dogs can understand finger pointing even though they don't have fingers, humans can understand some dog messages even though we don't have tails. The habit of co-operative communication can overcome physical barriers.
Thus, one understudied question is whether and when mammal intelligence changes on account of association with humans. Let us say that an indoor domestic dog or cat is freed from the need to hunt, protect herself, or raise offspring. Consequently, she enjoys a vastly increased life expectancy. Some such animals focus on status issues with respect to people and other dogs/cats, etc., generating layers of social complexity that are unlikely in a wilderness environment. She shows no progress toward human intelligence, but her human environment may determine how much canine or feline intelligence she lives to display.
"Feathered Primates" without Primate Brains?
Ravens can match or beat chimpanzees on some accepted tests of animal intelligence. Some researchers call crows "feathered primates." New Zealand crows' causal understanding (within limits) is said to rival that of 5-7 year old children. Or 7- to 10-year old children.
Some New Caledonian crows can use three tools in succession to reach food, and can also enact Aesop's fable by dropping stones into a jar of water till floating food rises.
It's not just crows. Pigeons' ability with numbers up to nine is "indistinguishable from that displayed by monkeys." Even the intelligence of the chicken "startles," according to Scientific American ("communication skills on par with those of some primates").
But how do we understand bird intelligence, given that bird brains show significant differences from mammal brains? And we can hardly fall back on common ancestry.
Language ability is an uncertain guide. Some birds are popularly held to be intelligent because they can imitate the human voice. This ability may be related to structural features of those bird species' brains:
In addition to having defined centers in the brain that control vocal learning called "cores," parrots have what the scientists call "shells," or outer rings, which are also involved in vocal learning.
That "shell" structure may be related to some parrots' ability to dance to music as well. But these birds probably don't know what they are saying or doing apart from the fact that, like the bicycling cockatoo, they are typically rewarded for doing it.
Alex the parrot (1976-2007), possibly the most famous "intelligent bird" personality, could use human language to communicate needs. However, he had only typical parrot needs. Alex was not achieving more human-like intelligence--as his researcher and patron Irene Pepperberg acknowledged:
"I avoid the language issue," she said. "I'm not making claims. His behavior gets more and more advanced, but I don't believe years from now you could interview him." She continued: "What little syntax he has is very simplistic. Language is what you and I are doing, an incredibly complex form of communication."
Put another way, if an intelligent dog had "vocal cords" (a syrinx) like a parrot, he could tell a human in words that he needs to go outside or have his water dish refilled. But he does not go on to express interest in things that do not naturally concern a dog.
One interesting thing we learned recently about smart crows is that they don't depend much on learning from each other (social learning):
Logan and colleagues found that the crows don't imitate or copy actions at all. "So there goes that theory," she said. ...
Even if one crow is at an apparatus and tries unsuccessfully to open the door, if he or she sees another crow on the second apparatus actually solving the problem correctly, the first crow doesn't use that information. "The social learning attracts them to a particular object and then they solve it through trial and error learning after that," Logan said.
The crows' pattern of learning seems different from human learning, and may be related to an inability to grasp or convey abstract information. Which brings us to the recent claim that crows fear death because many crows purposefully avoid places where other crows have died:
And this fear of a potential deadly situation stays with them. Even six weeks later more than a third of 65 pairs of crows continued to respond this way.
But, like the claim that chimpanzees mourn their dead, this one is founded on a misunderstanding: "Death" -- unlike danger or loss, which are experienced viscerally -- is a pure abstraction, like the number 23. An intelligent life form must understand not merely nature but the nature of nature to know what "death" means.
So we come to a culturally unexpected conclusion: Bird intelligence is a respectable competitor on a continuum with primate intelligence. But, like theirs, it is on a different track from that of humans.
Then There Are Those Cold-Blooded Reptiles and Fish...
A number of recent marketing strategies promise sales through appealing to a customer's "self-centered" reptilian brain. But that piece of business folk wisdom is based on a myth:
It is the idea that we have three brains: a reptilian one, the paleomammalian one and the mammalian one. The story goes that these were acquired one after another during evolution. The details differ with the writer. But it is all a myth based on an idea from the '70s of Paul MacLean which he republished in 1990. Over the years in has been popularized by Sagan and Koestler among others.
The brain is hardly so simple. Reptiles lack certain brain structures found in mammals, but like birds they sometimes use the ones they have for purposes that apparently display intelligence: Crocodilians (alligators and crocodiles) are reported to use sticks as decoys, play, and work in teams. Tortoises may well be smarter than once believed, though here we rely mainly on anecdotes, not formal studies, for now.
Even fish have shown signs of what seems like intelligence. We are told that pairs of rabbitfishes "cooperate and support each other while feeding":
While such behaviour has been documented for highly social birds and mammals, it has previously been believed to be impossible for fishes. ... "We found that rabbitfish pairs coordinate their vigilance activity quite strictly, thereby providing safety for their foraging partner," says Dr Simon Brandl from the ARC Centre of Excellence for Coral Reef Studies.
Why don't reptiles and fish appear intelligent? Here is a possible clue: Anole lizards were found as capable as tits (birds) in a problem-solving test for a food reward. But the anoles, being exothermic, don't need much food -- which hinders research. When reptiles and fish need to solve problems, they often use the brain structures available to them quite effectively. The rest of the time they may be comfortably inert. If so, the relationship between brain structure and intelligence is more complex than we have supposed.
Factors That May Promote Intelligence in Vertebrates
We have seen that, while brain structure is not the absolute limitation once supposed, cold-bloodedness (exothermic metabolism) may reduce the need for intelligence without actually preventing it. Conversely, living with humans may promote intelligence by creating systematic rewards for achievement. Nature, it is true, rewards intelligence, but not systematically, like a dedicated trainer seeking a response. So there are rough general trends in intelligence, as in evolution, but they appear to be patterns, not laws.
Do the patterns relate in some way to anatomy? Can we say, for example, that intelligence requires a multicellular life form that has a spinal column and a brain? What can the vast world of invertebrates tell us about that?
Denyse O'Leary December 22, 2015 3:27 AM
Much research on animal minds is rooted in Darwinian naturalist assumptions -- a long slow continuum of intelligence from somewhere just north of cytoplasm to humans. These assumptions may have set us back. First, just being a life form includes a drive to survive and an ability to adapt for that purpose, which we do not find in rocks. Many life forms can also communicate for those purposes. But, so far as we know, they lack consciousness or sentience, the ability to feel things. There is no slow ascent; there is a steep cliff.
At the other end of the spectrum are apes, who belong to the same order of life as ourselves. Discussions of their intelligence often assume that they are entering a "Stone Age" (such as the Lascaux cave artists lived in, 20 000 years ago). However, while apes tend to be more intelligent than most mammals, they are not becoming like humans. And smart birds give apes serious competition, when tested.
So what can we learn from other vertebrate life forms, forms that show intelligence but are not closely related to us, do not seem much like us, and are not apparently heading in our direction?
Can Animal Mind Be Explained by "Instinct"?
At one time, it was supposed that most animals were simply born with instincts about what to do. The term is not used much now because it mainly meant that we do not know the source of the animal's information. We are now learning many of these sources.
We have recently discovered, for example, that migrating birds can use the mineral magnetite, embedded above their beaks, to use Earth's magnetic fields for direction.
We learned in recent decades how young birds "know" that they should follow their mother: As Spark Notes explains, following the work of animal behaviorist Konrad Lorenz (1903-1989):
Johnson and Bolhuis identified two independent neural systems that control filial imprinting in precocial birds. Newly hatched chicks will follow almost anything that has eyes and moves. After the chick follows something, another part of the brain, analogous to the frontal cortex, recognizes and imprints on the individual being followed. These mechanisms are independent. There is an instinct for chicks to follow, and then they learn what they are following.
But the "follow her" system is not strictly genetic:
It might seem odd that being able to identify and follow a mother does not have a genetic mechanism. Yet with a neural rather than genetic mechanism, the chick gains flexibility that might help in survival. If a chick's mother dies, the chick can then be adopted by another family member or conspecific.
Yes indeed. Famously, the young bird may follow a psychology student, a stick, or a cat, with varying results. Bird rescuers use hand puppets of bird faces when caring for nestlings, to return them later to a natural setting. Clearly, not all we need to know about an organism is in its genes.
How then does the male weaverbird know how to build a nest? That's apparently not simply a genetic program either; the birds must learn some of the techniques by experience.
Genetics, neural networks, and experience all make animal learning much more complex and information-rich than the concept of "instinct" implied. But we are not yet in the realm of "intelligence." The migrating and nest-building birds access existing solutions to longstanding problems; they do not come up with new ones.
Both birds and mammals can learn to solve new problems presented to them. Let's look at some recent finds in mammals first, bearing in mind that we have only really begun to look at their intelligence seriously. It is early days yet, so some sketchiness is inevitable.
Mammals' Unexpected Intelligence
We find intelligence where we did not expect it. Pigs, for example (despite their reputation), are "socially complex as other intelligent mammals, including primates" (Natural History Magazine). That is surprising because pigs don't usually form close relationships with humans, as dogs do. And hog farming operations don't encourage intelligence.
We know more about the intelligent animal we are close to. In terms of communication, horses' surprisingly varied facial expressions are more similar to those of humans on one measurement than those of chimps are:
The Equine Facial Action Coding System (EquiFACS), as devised by the Sussex team in collaboration with researchers at the University of Portsmouth and Duquesne University, identified 17 "action units" (discrete facial movements) in horses. This compares with 27 in humans, 13 in chimps and 16 in dogs.
That might account for the human-horse bond (there seems no similar chimp-horse bond).
Horses', dogs', and cats' tail communications are also easy to read (they are intended to be). So just as dogs can understand finger pointing even though they don't have fingers, humans can understand some dog messages even though we don't have tails. The habit of co-operative communication can overcome physical barriers.
Thus, one understudied question is whether and when mammal intelligence changes on account of association with humans. Let us say that an indoor domestic dog or cat is freed from the need to hunt, protect herself, or raise offspring. Consequently, she enjoys a vastly increased life expectancy. Some such animals focus on status issues with respect to people and other dogs/cats, etc., generating layers of social complexity that are unlikely in a wilderness environment. She shows no progress toward human intelligence, but her human environment may determine how much canine or feline intelligence she lives to display.
"Feathered Primates" without Primate Brains?
Ravens can match or beat chimpanzees on some accepted tests of animal intelligence. Some researchers call crows "feathered primates." New Zealand crows' causal understanding (within limits) is said to rival that of 5-7 year old children. Or 7- to 10-year old children.
Some New Caledonian crows can use three tools in succession to reach food, and can also enact Aesop's fable by dropping stones into a jar of water till floating food rises.
It's not just crows. Pigeons' ability with numbers up to nine is "indistinguishable from that displayed by monkeys." Even the intelligence of the chicken "startles," according to Scientific American ("communication skills on par with those of some primates").
But how do we understand bird intelligence, given that bird brains show significant differences from mammal brains? And we can hardly fall back on common ancestry.
Language ability is an uncertain guide. Some birds are popularly held to be intelligent because they can imitate the human voice. This ability may be related to structural features of those bird species' brains:
In addition to having defined centers in the brain that control vocal learning called "cores," parrots have what the scientists call "shells," or outer rings, which are also involved in vocal learning.
That "shell" structure may be related to some parrots' ability to dance to music as well. But these birds probably don't know what they are saying or doing apart from the fact that, like the bicycling cockatoo, they are typically rewarded for doing it.
Alex the parrot (1976-2007), possibly the most famous "intelligent bird" personality, could use human language to communicate needs. However, he had only typical parrot needs. Alex was not achieving more human-like intelligence--as his researcher and patron Irene Pepperberg acknowledged:
"I avoid the language issue," she said. "I'm not making claims. His behavior gets more and more advanced, but I don't believe years from now you could interview him." She continued: "What little syntax he has is very simplistic. Language is what you and I are doing, an incredibly complex form of communication."
Put another way, if an intelligent dog had "vocal cords" (a syrinx) like a parrot, he could tell a human in words that he needs to go outside or have his water dish refilled. But he does not go on to express interest in things that do not naturally concern a dog.
One interesting thing we learned recently about smart crows is that they don't depend much on learning from each other (social learning):
Logan and colleagues found that the crows don't imitate or copy actions at all. "So there goes that theory," she said. ...
Even if one crow is at an apparatus and tries unsuccessfully to open the door, if he or she sees another crow on the second apparatus actually solving the problem correctly, the first crow doesn't use that information. "The social learning attracts them to a particular object and then they solve it through trial and error learning after that," Logan said.
The crows' pattern of learning seems different from human learning, and may be related to an inability to grasp or convey abstract information. Which brings us to the recent claim that crows fear death because many crows purposefully avoid places where other crows have died:
And this fear of a potential deadly situation stays with them. Even six weeks later more than a third of 65 pairs of crows continued to respond this way.
But, like the claim that chimpanzees mourn their dead, this one is founded on a misunderstanding: "Death" -- unlike danger or loss, which are experienced viscerally -- is a pure abstraction, like the number 23. An intelligent life form must understand not merely nature but the nature of nature to know what "death" means.
So we come to a culturally unexpected conclusion: Bird intelligence is a respectable competitor on a continuum with primate intelligence. But, like theirs, it is on a different track from that of humans.
Then There Are Those Cold-Blooded Reptiles and Fish...
A number of recent marketing strategies promise sales through appealing to a customer's "self-centered" reptilian brain. But that piece of business folk wisdom is based on a myth:
It is the idea that we have three brains: a reptilian one, the paleomammalian one and the mammalian one. The story goes that these were acquired one after another during evolution. The details differ with the writer. But it is all a myth based on an idea from the '70s of Paul MacLean which he republished in 1990. Over the years in has been popularized by Sagan and Koestler among others.
The brain is hardly so simple. Reptiles lack certain brain structures found in mammals, but like birds they sometimes use the ones they have for purposes that apparently display intelligence: Crocodilians (alligators and crocodiles) are reported to use sticks as decoys, play, and work in teams. Tortoises may well be smarter than once believed, though here we rely mainly on anecdotes, not formal studies, for now.
Even fish have shown signs of what seems like intelligence. We are told that pairs of rabbitfishes "cooperate and support each other while feeding":
While such behaviour has been documented for highly social birds and mammals, it has previously been believed to be impossible for fishes. ... "We found that rabbitfish pairs coordinate their vigilance activity quite strictly, thereby providing safety for their foraging partner," says Dr Simon Brandl from the ARC Centre of Excellence for Coral Reef Studies.
Why don't reptiles and fish appear intelligent? Here is a possible clue: Anole lizards were found as capable as tits (birds) in a problem-solving test for a food reward. But the anoles, being exothermic, don't need much food -- which hinders research. When reptiles and fish need to solve problems, they often use the brain structures available to them quite effectively. The rest of the time they may be comfortably inert. If so, the relationship between brain structure and intelligence is more complex than we have supposed.
Factors That May Promote Intelligence in Vertebrates
We have seen that, while brain structure is not the absolute limitation once supposed, cold-bloodedness (exothermic metabolism) may reduce the need for intelligence without actually preventing it. Conversely, living with humans may promote intelligence by creating systematic rewards for achievement. Nature, it is true, rewards intelligence, but not systematically, like a dedicated trainer seeking a response. So there are rough general trends in intelligence, as in evolution, but they appear to be patterns, not laws.
Do the patterns relate in some way to anatomy? Can we say, for example, that intelligence requires a multicellular life form that has a spinal column and a brain? What can the vast world of invertebrates tell us about that?
Subscribe to:
Posts (Atom)