100 Days Blog

Day 059 - Evolutionary arms races

Submitted by Sam on 18 July, 2011 - 23:56

Co-evolution can engender an arms race between competing genes, as predators adapt to better catch their prey and prey correspondingly counter-adapt to better evade their predators. Each genetic lineage 'races' against the other, progressively improving and counter-improving over many generations, always trying (in so far as a gene that produces a particular protein which has some effect on behaviour or morphology can be said to 'try') to out-compete the rival. A common-sense example is the asymmetric genetic arms races between cheetahs and gazelles, where one evolves adaptations which tend to make it better at chasing whilst the other consequently co-evolves adaptations which favour a greater evasive ability. A symmetric arms race occurs in the height of trees in a forest, where the selection pressure for access to light tends to cause trees to grow taller, amongst other adaptations.

These intuitive examples are inter-species arms races, but there are subtler, more insidious intra-species conflicts, even between the sexes. One interesting example is provided by the powerful effect a male canary's birdsong has on a female canary's reproductive behaviour; she initiates nest-building behaviour and increases the size of her ovary when she hears it. This has been shown to happen even when the stimulus is only a recording of a male canary's song, suggesting that the song acts like a drug, manipulating her to be primed for the male's advances. One reading of this phenomenon is that the male has 'won' an arms race, manipulating the nervous system of the female to his own ends, using his song to create an electrochemical pattern in her brain, which in turn stimulates her pituitary to synthesize the required hormones to bring her into an appropriate reproductive condition. Alternatively, the female may exhibit this behaviour as a counter-adaptation which requires the male to give an exhaustive performance of his song before she is willing to mate, signalling that he is a robust partner and one that is willing to commit time and energy to the relationship (having already spent time and energy wooing with a song).

Day 058 - Choosing a mate

Submitted by Sam on 17 July, 2011 - 17:48

As far as the male's genes are concerned, the best strategy for propagation is to mate with as many females as possible, always abandoning one for another and forcing them to bring up his children themselves, affording him more time to mate by never having to expend resources in childcare. This strategy is only possible if the population of females permits it; if the majority of females choose instead to withhold mating until the male completes costly courtship tasks, he would never be rewarded for desertion as he would always face a repeated series of time and energy consuming trials before being able to mate again. Conversely, this strategy is unstable if there are many loose females in a population, as a male would always have someone to run to having deserted his mate.

Female elephant seals permit the male strategy 'mate with and abandon as many females as possible'. However, their social structure ensures that only one bull mates with the harem in any season, leaving the rest of the males with only opportunistic copulations when he is otherwise engaged. The other males will only stand a chance of spreading their genes if they can defeat this alpha bull in fighting, and become the leader of the harem themselves.

This subjugation by a single dominant male may have originally arisen through discrimination on the part of the females, who could have refused to copulate with all but the 'best' males, those who bore visual indications of carrying 'good' genes, that is, genes which confer survival advantages, like strong muscles or big, sharp teeth. By breeding only with these select males, females would increase the survival prospects of their children, and simultaneously benefit her own genes and creating a selection pressure for males with visual indications of health and strength, or promoting sexual attractiveness itself. As the females will be judging the males on the same criteria, only a few of them will be selected for breeding, and these will be those with characteristics that equip them well for fighting, or defending their mating rights from rival males: the harem structure is born.

Although not obviously the case in elephant seal populations, females selecting mates on the basis of physical indicators of survival prospects can create a selection pressure for males with these attractive qualities, whether they are actually useful for physical survival or not. The elaborate tails of the male birds of paradise, for instance, may have evolved this way: longer tails may initially have carried some indication of a male's health and ability to find food, and thus the suitability of his genes to create strong, healthy children. After generations of selection for bigger, more extravagant tails, males may actually have been left with tails so large that they became a physical disadvantage, possessing them merely because they were attractive to females, and no longer functioning as a truthful indication of physical prowess.

This is a self-reinforcing cycle, because if long tails were considered by the majority of the female population as desirable, any female mating with a male with a shorter tail would stand little chance of her own son being regarded as a potential mate, and therefore she would disadvantage the spread of her own genes. In this way, genes set a standard for sexual attractiveness which it pays to stick to.

Day 057 - Differences between the sexes

Submitted by Sam on 17 July, 2011 - 00:11

Relative to each other, males tend to produce small sex cells and females tend to produce large sex cells; sperm are small and many, eggs are large and few. Apart from being one of the most convenient ways of distinguishing males from females in both animals and plants, this difference in gamete size is symptomatic of a fundamental asymmetry between the sexes. Each sex has to establish a compromise between time and energy spent raising children to ensure they are healthy and that their genes will continue to propagate, and time and energy spent fighting with rivals for mates and other resources, to maximize the number of children containing these genes. If one sex is to settle on a different balance between these two opposing needs, perhaps initiated by a difference in resource allocation so slight that it arose at random, it is likely to be selected for so that parents who are successful in fighting breed fighters who benefit their genes more from fighting more than parenting, and vice versa. This selection will naturally lead to an escalation in differences until one sex achieves reproductive success primarily through investing in parental behaviour whilst the other ensures their reproductive success primarily through fighting rivals to mate with the most.

This difference is apparent right from conception, where the unequal size of the male and female gametes informs an uneven distribution of resource investment in the child in the earliest possible stages – the male has invested less than his fair share of half of the necessary energy to produce the fertilized egg, whilst the female has had to invest comparatively heavily in her large, energy-rich egg. Whilst both parents have contributed an even number of genes to the child, their sex cells contribute an uneven share of resources.

Unconsciously impelled by their genes, parents try to exploit each other to invest more than their fair share in their child. It is advantageous, from their gene's point of view, to let the other half do all the work and take the burden of child rearing, allowing the shirking parent more time and energy to pursue other sexual partners, and thereby spread more copies of their genes. The basic characteristics of the male sex cells reflect this disparity, as many small, fast sperm allow him to potentially sire many more children than a single female could raise. As eggs are larger than sperm, the female is evolutionarily exploited at the earliest stage, and she is placed in the position where she must invest more in the child's rearing than the male. In mammals this is especially pronounced, as the mother must incubate the foetus herself, feeding it from her own body, investing more than the father and therefore becoming more 'committed' (resource-wise) to the child. She stands to lose much more from the loss of the child than the father does, and is therefore much less likely to abandon the child when it is born. The father, who has much less to lose if he abandons mother and child, will have more time and energy to fight for another mate and beget more children if he does so. There is a battle between the parents as to who deserts first, and who can get away with the least amount of parental investment.

Perhaps the only option the female has to redress this unfavourably weighted balance is to refuse to mate with the male until he pays a high price for her prized commodity, her large, nutritious egg for his child, making him invest time and energy in her and the unborn child before conception. Such pre-copulation investment take the form of courtship rituals, and include nest building and offerings of substantial amounts of food. By extracting a price for mating, the female is in a position to tie the male in through his investment, perhaps decreasing the chances that he will desert after mating, and instead continue to devote himself to the well-being of his child.

Day 056 - Favourite children

Submitted by Sam on 15 July, 2011 - 20:10

As far as proportions of genes are concerned, there should be no reason why a parent should invest more time and energy in one offspring rather than another, as each parent has the same relatedness to their children. Although it would seem that a parent might best serve the proliferation of their genes by investing equally in all of their children, there are invariably differences in life-expectancy between a range of offspring, and some are better bets than others. Despite containing the same proportion of their parent's genetic material as their siblings, undersized or otherwise disadvantaged children (or runts of litters) have a much lower life expectancy, and so would require a greater than normal parental investment just to be given an equal chance as the more advantaged children. In such cases, it may be worthwhile for the mother to invest more in her other children and refuse to feed the weakest child, instead spreading its allocation of 'parental investment' to the other children. Following this line of reasoning to its logical conclusion suggests that the optimum strategy might in fact involve eating the runt, or feeding it to its siblings in order to reclaim some of the lost investment in energy.

This imbalance creates a tension between the generations, as children endeavour to manipulate their parents in order to receive more than their fair share of investment, whilst parents must endeavour to identify such exploitation and allocate resources in the most (genetically speaking) efficient way possible.

Day 055 - Selfish and selfless

Submitted by Sam on 15 July, 2011 - 00:04

Whilst a gene for saving close relatives from death (even at the expense of your own life) could theoretically spread in the gene pool, it could only be successful if its bearer is actually able to identify its close relatives, which is not necessarily an easy task. One way for an organism to recognize its kin with some degree of reliability would be to remember other members of the species who share a physical resemblance to them. A gene encoding the behavioural equivalent of 'be nice to those who look like you as they might be your relations' would generate the kind of altruistic behaviour that would mutually support the spread of selfish genes shared through different bodies.

In some species that stay in small groups or who do not move around much, there is a good chance that any member is closely related to another, and so genes which tend to promote altruistic behaviour towards any member of the same species may tend to spread through the gene pool, as any possessor would be more likely to look out for other possessors than not. Dawkins offers the example of a male baboon defending its troop from predators, risking its life to protect the genes which are statistically probable to be invested in other members of the group, which may contain many close relations.

However, no matter how altruistic the behaviour encoded by genes is, it can never be as strong as the encoded propensity towards individual selfishness, simply because your genes can only ever be completely certain of your own individual identity, whilst altruistic genes can never be sure to completely recognize close kin, as they are susceptible to errors in classification and, may mistake perfect strangers for close kin.

Day 054 - Dying to save your family

Submitted by Sam on 13 July, 2011 - 21:34

The most extraordinary thing about genes is that they can programme their survival machines to behave in seemingly altruistic ways, even to the point of that machine's self-destruction. A particularly extreme example would be an individual sacrificing itself in order to save the lives of others. Dawkins' selfish gene theory gives an insight into how such an apparently altruistic trait could develop to the gene's advantage by showing that the closer the genetic relationship between two individuals the more sense it makes for them to behave altruistically towards each other, as each will be statistically likely to be carrying a good proportion of the other's genes.

As close relatives have a greater average chance of sharing genes, it could conceivably benefit the spread of a particular individual's genes if it died in order to rescue ten close relatives from drowning, for instance, as it would be likely that whilst one copy of the 'kin-altruism' gene would be lost in the process, many more may be preserved through the apparently selfless act, thereby increasing the 'save close relatives from drowning, even if it kills me' gene's prevalence in the gene pool. It should be cautioned that there is of course no such gene, merely a gene which, when present and expressed in the complex matrix of other genes in a body, induces a greater than usual tendency to save people from drowning than if it were not present or expressed.

Crucial to an understanding of this example is the fact that whilst such a gene may be rare in the wider population, it is likely to be common within a family. There is a 50% chance that your siblings contain the same particular rare gene that you do, as they would have received the gene either from your father or your mother. This means that the 'relatedness' between two siblings is ½, since on average, half of the genes each has will also be possessed by the other. The relatedness between a parent and a child will always be exactly half (barring any mutation), because the process of meiotic division ensures that 50% of a child's genes are received from its father's sperm and 50% are received from its mother's egg. An individual's relatedness to itself is 1, because it can be sure it has 100% of its own genes.

Through this index of relatedness, it can be shown that a gene for suicidally saving five of your cousins (who each have a 1/8th relatedness to you, producing a total relatedness of 5/8ths) will not spread throughout the gene pool, whereas a gene for saving five brothers (each with a ½ relatedness, totalling a 2 ½ relatedness), or ten cousins (1 2/8ths) would spread throughout the population. This is because a suicidal altruistic gene only has to be successful if it can save a minimum total relatedness of more than 1, because this is the total (arbitrary) 'value' of the genes which will be lost through the altruistic suicide. If a gene can, on average, save more than a minimum of two brothers or sisters or parents, more than four uncles, aunts, grandparents (and so on), then it will persist in the bodies of the saved individuals with a frequency great enough to compensate the loss of the altruistic individual.

Day 053 - What survival machines want and what they get

Submitted by Sam on 12 July, 2011 - 22:35

Bodies are the vehicles of survival for genes, which programme proteins to make systems which will live long enough to reproduce, ensuring that the genes themselves continue to replicate. Whilst genes obviously have no conscious purpose or desire to replicate, it is their natural propensity to create organisms that do exhibit such behavioural characterstics, whether consciously or otherwise. Following in Dawkins' tradition then, it is convenient to refer to the effects of genes with the compressed language of intention, imagining what they might 'want' as a short-hand way of referring to the behaviour their interrelated effects tend to produce in their host organisms to benefit their genes' own propagation. What a gene 'wants', in short, is to ensure its own survival, by replicating as widely as possible throughout a population.

To this end, genes are 'selfish', largely seeing other survival machines either as impediments or as resources to be exploited. The important exception to this indifference is when those other machines are close relatives to the first, when they will likely be vehicles for the same genes as it carries itself. If they are not genetically closely related, other survival machines represent a threat, potentially competing for mating partners or other resources. The logical policy to adopt against this threat might seem to be for each organism to murder all of its rivals, and perhaps even then eat them for food. However, this policy does not necessarily always work, as removing one rival from the complex system of many rivals may benefit yet another rival more than it benefits oneself.

What tends to happen therefore, is that various survival machines adopt various pre-programmed behavioural policies, perhaps of the form 'if rival is smaller, attack; otherwise flee'. Some strategies be more effective than others, and will therefore tend to spread throughout the population. What is crucial is that the behaviour of each individual depends largely on what the majority of the other members of the group are doing – if the majority of other genes are encoding the behavioural pattern 'if rival is smaller, run away; if larger, attack', then clearly a selective advantage is shown to those that exhibit the reverse trait. After a period of oscillation, an evolutionarily stable strategy will be adopted; a pre-programmed behavioural policy which cannot be bettered by any alternative strategy if most members of the group follow it. It may not necessarily represent the optimum efficiency for the group were they to conspire together to create a perfect strategy, but instead gains its persistence from being 'immune to treachery from within' – once such a strategy has evolved through the testing of many other forms, a deviant individual cannot, on average, out-perform it. If it can, and has evolved a newly successful strategy, then another period of oscillation will occur until this is selected as the new evolutionarily stable strategy for the group to adopt.

Day 052 - Where bodies came from

Submitted by Sam on 11 July, 2011 - 23:07

Intelligence arises from a society of unintelligent agents working together in very complex ways, emerging from a group of neurons that can, individually, be represented algorithmically. But where do neurons themselves come from? Where do all biological structures come from? How can evolution produce something so complex as an electrically-gated logic circuit from raw materials like carbon and hydrogen, insensibly pulling together groups of molecules to create structures as intricate as the organic computers that power the human brain?

In his seminal book The Selfish Gene, Richard Dawkins extends Darwin's theory of natural selection backwards in time to the inorganic precursors to life, lensing Darwin's principle of 'survival of the fittest' into the more general 'survival of the stable' to show how complex yet stable molecules 'evolve' from their constituent elements.

Chemists have shown how a sea of simple substances, like those that would have been found in the earliest environments on earth, can be stimulated with a source of energy, such as sparks (representing primordial lightning) or ultraviolet light, to produce molecules more complex than those originally present in the mix. Typically, after a few weeks, a 'soup' containing amino acids is created. Amino acids are the basic components of proteins, which are in turn the basic components of biological organisms. Dawkins saw that the earliest form of natural selection was therefore a selection of such stable molecular forms against a rejection of unstable ones, which would rapidly and automatically degenerate to be replaced by more robust forms.

Such a 'primeval soup' is believed to have constituted the seas thousands of millions of years ago, where clumps of organic molecules could become locally concentrated, perhaps combining into even larger molecules under the influence of energy from the sun or lightning. Through countless iterations of this process, Dawkins theorized that a singularly remarkable molecule was formed by accident – a molecule able to create copies of itself which he termed tht replicator.

A replicator could have acted as a mould or a template built from smaller building block molecules derived from the abundant soup, arranged in such a way that each building block had an affinity for its own kind. With this chemical propensity to draw together and bond, the building blocks would automatically join together into a sequence that would mimic (or inversely mirror) the shape of the replicator itself. If the chain ever split, there would be two replicators, and they would spread rapidly through the soup of building-blocks, making further copies of themselves.

This process would continue until the smaller molecular components in the soup became a 'contested' resource, placing a selection pressure on replicators which would favour any which might happen to form that used even larger molecules as building blocks. Non-identical replicators may have been created through copying errors in their 'parent' replicators, and these cumulative mistakes could have created molecules even more stable in the new environment than the old replicators. Perhaps new replicators formed by accident which were able to cannibalistically break down other replicators, decreasing their stabillity in order to obtain 'food' to fuel their own replication. In response to these 'proto-carnivores', other replicators may have been selectively favoured if their copying mistakes afforded them chemical or physical protection from their rivals, perhaps through a physical wall of protein which shielded them from chemical assimilation.

Dawkins shows that the replicators which would survive in this soup would be the ones that built 'survival machines' for themselves to live in, which would gradually get stronger and more elaborate as the competition for resources grew closer, and the competition itself grew more advanced. The culmination of these survival machines, and the marvellous conclusion to Dawkins' argument, is that genes are descended from the original replicators, and we are their survival machines.

Day 051 - Defining things

Submitted by Sam on 10 July, 2011 - 23:22

In order to have a notion of a “thing” we first have to be able to distinguish a group of properties which seem to stay the same whilst other things change, or that change in ways that we can predict. Our 'thing-recognition' processes have evolved to isolate an integral constellation of properties to accurately define an entity, whether a spoken word or a physical object or anything that can have a name. This ability is so elementary to us that we hardly ever consider how marvellously complex it is, given that we never see (or hear, or feel, or taste) the same thing in exactly the same way twice – invariably we always experience it from a slightly different perspective, perhaps against a different background, from a higher perspective, or under a different shade of light. Whilst our ability to differentiate things seems simple because it is 'second nature' to us, this is only because we are unconscious of the great network of processes in our brain that make the fantastically intricate computations that allow us to make such distinctions possible.

Even the most advanced robots today lack the visual object-recognition capabilities that a two year-old child possesses, and fail to attain the language capabilities of a four year old child. A two-year old child can recognize classes of objects, and can classify a black shoe as a shoe, despite never having seen a black shoe of that exact style, colour or size before, but yet our robots cannot reliably classify because we have yet to be able to replicate the complex object-identification processes in our brains. A four year-old child can understand language in noisy environments and in a variety of accents, whilst our voice-recognition algorithms break down under such conditions.

Humans are able to discard a great deal of sensory information in order to perceive only what is most essential to each scene – without this ability we would be unable to learn because our memories would never sufficiently correlate with current appearances.

Day 050 - Expressing yourself

Submitted by Sam on 10 July, 2011 - 01:36

We can never verbally express exactly what we are thinking at any given moment, because to do so would require articulating the states of agencies that we are not actually conscious of, let alone able to describe using language agencies. 'What I am currently thinking' can therefore only ever be an expression of higher-level agencies, and therefore only a partial indication of the global brain state, leaving out a description of its non-verbal emotions and thought-processes. In endeavouring to translate the states of the brain that we are consciously aware of into language, there is also an inevitable time delay, implying that any expression of the 'current' state is either an anticipation of what higher-level agencies will be doing by the time the description is vocalized, or a reflection as to what they were doing 'just now'. By the time you have expressed what you were thinking, your state of mind has changed, and new thoughts have been formed as a result of this attempt at introspection and expression.

The same problems occur when we try to express an idea to someone else; we often end up not entirely saying what we 'meant'. This is because if the idea pre-exists in our brain as some kind of structure, it is not necessarily going to be a definite, fixed structure that language agents can easily reformulate into an easily transmittable description – not least because some parts of the idea may be reliant on interactions with a rapidly changing network of agencies whose subtle interactions are not accessible to conscious thought or linguistic expression. In order to try to express the idea then, the language agencies in the brain must hypothesize about the states of these linguistically-inaccessible states, attempting to reformulate them into words. This process inevitably oversimplifies the transmission of the mental state necessary to recreate the idea in its true essence, perhaps leading to a loss of nuance and full comprehension.

The best descriptions work by using words which decompress in the mind of the reader or the listener, activating a whole series of networks and associations, both conscious and unconscious, which expand into a recreation of the original 'meaning' in its purest sense, activating both the language areas associated with the original idea and those attendant agents which were not expressed verbally. Through the reassembly of the words using the listener or reader's own personal lexicon of inferences and definitions, a similar structure to that of the original idea can be rebuilt in the 'receiving' mind, even though it is only ever a representation or reformulation of the original neuronal activity that constituted the first idea.

Syndicate content
Drupal theme by Kiwi Themes.