By Tom Veatch
Functional Logic of Evolution
Evolutionary logic rests on tautologisms: Statements that say nothing, that are true by definition. (So you can't deny them, or argue with them; inexorable, they hold us all helpless. Therefore submit, O brave heart, and see what you may learn!)
If an organism survives then it survives.
If it reproduces then it reproduces.
If it dies then it dies. (But, why and how death itself evolved is a non-tautological question.)
If jointly-functional gene groups half-subset and combine compatibly between individuals creating a new individual, then sexual reproduction occurs.
If there are more of one kind then there are more of that kind.
The past has traction into the future by means of information existing or living in the past that continues to exist or live in the future.
Survival and reproduction are required for past life to continue into the future.
If the organism dies without reproducing then it, and its particular information, dies.
The world in which life exists and continues into the indefinite future is a world of sufficient resources, energy, and time or stability, for life, at least for now.
A group of essences or useable informational units, mutually compatible and jointly life-supporting, undergoing mutual recombination and replication, depend on each other to survive in essence into the indefinite future. They are like mutually substitutable sentences in a story, each organism being a single story or sequence of selections from the substituteable sets; the living population is the set of life stories currently selected and the species itself is temporarily statistically stable unit, or you might say, the abstraction, which is the commonality across the set of all those stories or narrative subselections, defined in biochemistry by the actual choice points and actual choices chosen there, and defined in bitter logic by their compatibility with and enhancement of mutual or shared survival. Without successful cooperation among these essences, the many choices cannot individually survive, so mutual compatibility is a filtering requirement on the choices and their interactions.
If a combination of essences is compatible in a contemporaneous environment with multiplication, and multiplies, then that combination multiplies there and then.
I haven't said anything yet. Tautology. Like 1=1.
If a life form does not have systems in it which tend to, nay, which reliably, make it survive and reproduce in its environment, it won't (let me sometimes write "S&R" for "survive and reproduce"). Therefore effective S&R mechanisms are necessary for an organism and will be present in any evolved species.
Directions and path-sequences being innumerable, and pathways compatible with survival being few and limited in character, it follows that if a living system doesn't (let's personalize it, if you don't) aim at your target, then you are not likely to hit it. These mechanisms may seem designed for an intended purpose only because the link between mechanism and its associated outcome must be rather tight in order to be functional, that is reliably tending toward survival or reproduction or the observed subsidiary outcome. In cases where a survival-connected outcome is only loosely predictable from the whole mechanism, the carriers of the mechanism will only loosely survive in one generation, and over many generations that total probability will approach zero. Functional logic is harsh and controlling, when the tautologies of evolution impose it on life.
From this follows the prioritization of survival before reproduction where separable.
If survival depends on multiple subordinate outcomes, and each subordinate outcome has an evolved, effective mechanism to satisfy it, then the competitive prioritization of all the mechanisms (for their respective outcomes) must be done by a competent meta-mechanism, which is able to make and enforce a detailed rank ordering according to system state or circumstances, so as to orchestrate the few survivable and reproducing outcomes reliably.
Bits in Evolution
Maslow's hierarchy of needs, the Hindu chakra system, the Freudian plural id, the hormonally-governed subpersonalities, the at least hypothalamic system for choosing which subpersonality must be in charge when: these are perhaps expressions or homomorphisms of a single, of one and the same, functional logic and prioritization constraint set imposed on complex life by the tautologies of life. Of course survival first. Respiration, temperature, rest/sleep, sustenance, have their nonlinear hierarchy of systemic control, which may derive from the duration of time you can survive without that particular homeostatic requirement being met. The sooner you die without some requirement, the higher the priority and the more rigid the enforcement. This seems like tautology, but we now have mechanism linking to functional outcome: were it not so, were it instead that mechanisms were not linked to outcomes, then those fatal outcomes would fail to be defended against, and would occur, and those life forms will not last long, as the chances multiply. So long-term evolutionary survival statistically implies (over thousands or millions of generations it more and more categorically implies) strong and reliable, carefully calibratedly mutually prioritizing, mechanisms to defend against each of the various modes of practical failure, of death or non-reproduction.
Observe: Functional outcome is categorical and discrete. It can be expressed in hard yeses and noes, binary digits, bits, category-based logics, despite the physical continuity of every kind of dimension, measurement, control variable or physical Blob in this world at least where we are talking, far above the level of quantum chemistry. Because there are Outcomes: simple, discrete, dispositive facts. You live, or you die. You reproduce, or you don't. You turn left or turn right, and thus approach food or flee toxicity. In this way evolutionary outcomes or constraints are isomorphic to what comes to us as sets of linguistic categories: discrete, typically binary assertions; thus we may talk about the "choices" and "motivations" of an organism. In this way organisms are constrained to survive in a symbolically expressible world. Not a world of mere low level, continuous physicality, but a world of choice, decision, and action: of discriminating and motivated frames.
This is so not just at the level of conscious, linguistic, supposedly rational humanity, for which internal symbol representation and manipulation and, for example, discrete planning processes are axiomatic among cognitive scientists. It is even true at the level of paramecia, of single-celled mobile creatures.
Because if the life form doesn't have mechanisms that reliably produce surviving, reproducing outcomes, which are discrete outcomes, then it won't continue to evolve, but will drop out even from the very dustbin of history into the vacuum of nonexistence, as something that never even needed to have been tried as an experiment.
A scientist's job then includes the cognitive scientist's job.
The cognitive scientist, remember, creates categorical logic systems mapping between discrete informational inputs and outputs, using inferred internal "mental" representations that can be modelled using the computer metaphor with bits and memory and goals and actuators. Is that just the mental masturbation of overexcited grammarians? Or is it a necessary level of description of the functional logic of what actually ends up happening within the context of discrete, evolutionarily unavoidable and potentially bitter outcomes. I say the latter. Because if you can't get that logical system right in describing a biological system, then you don't understand why the thing is able to actually survive and reproduce, and more fundamentally even if you know the physics, the chemistry, the biochemistry, the cell biology, every detail of every subsystem, you still don't have any handle on why it survives or not: you don't understand it. System understanding depends on the cognitive scientist's modelling. Tell me I'm wrong.What are these discrete categories and the motivational frames of organismal existence, what the rank ordering of priority, what the logic controlling their relations and community of interest, and the communications or interactions they must have, if only in sharing to an internal Master Controller their several concerns and status reports or awarenesses for it to mediate or sort. (A scientist must also describe these discrete, cognitive behaviors within the functional logic that applies to them. For if the organism doesn't follow that logic, it will shortly be gone. This opens room for cognitive modeling of animal behavior, not just of humans or human language.)
For example, it seems that the driven-personality quality of people, as we are, is a quality equally driven and necessary in the least watery wiggler. If an organism isn't driven, and doesn't drive toward whatever motivations can be ascribed to it for its survival, it doesn't survive. The tautologies apply, and bye-bye.
Consider attention itself, consisting in the continual reassessing of priorities and of situational awareness tending to inform those priorities, would seem to be a highly valued early evolutionary achievement. For otherwise rapidly changing conditions would be fatal, and the Tautologies of Evolution apply. Similarly, integrated behavioral control, whereby a single target of action is aimed at, since one can hardly hit two targets at a time much less more, would also seem highly valued, evolutionary speaking. For if competing incompatible priorities are allowed to remain unresolved in a circumstance of potential action, no target may be attained, an outcome likely to be, or evidently, evolutionarily inferior to one where the most important single target was actually aimed at and acted towards or into being.
Mechanisms might be different for the same functional outcome, if new ones might be evolved that are as good or better, but the functional logic is the same, from paramecia to humans. Functionally one could say those mechanisms are the same system. They live in the same functional box, subject to the same logic. No efficient designer, neither Ockham nor hurried evolution itself, would fail to reuse the contents, by the way, so why we should think we are so different from worms with our special uniquely human characteristics of attention, spatial awareness, integrated action control (a.k.a. Will), I don't know.
I have argued for the tautologies of evolution imposing a symbolic, discrete logic on successfully evolved (all) organisms. From what may evolve, I turn next to when things may evolve, in what order.
Putting changes in sequence
We are given a descendant species with some set of novel characteristics N as compared with the shared set of characteristics Sh among the other species in the genus. Are we perhaps entirely unable to reason about the sequence by which the N characteristics evolved? There may be no archeological or DNA data on the question. But suppose it can be established that some feature N1 is logically prerequisite for another feature N2, then we can at least provisionally infer that N2 evolved after N1. Agreed? If you logically can't have N2 without N1, then N1 must have preceded N2.
Is this the limit of our powers? I don't think so.
Suppose further that N2 is easier to evolve in the circumstance where N1 exists: not impossible, but easier. May we make the parallel inference? Potentially Yes, I think so. The nature of evolution is such a grand statistical drift that statistical effects are very likely to prevail.
If a certain result associated with N2, such as a measureable mean, or a population frequency of a feature, etc., is, say, P% more likely to be reproduced or enhanced with N1 precedent and present than without, then the fact that generation after generation for tens and hundreds or thousands of generations has produced the effect, means that the full-sequence likelihood of it not being in that order is (100%-P%)^G with G the number of generations. If P is 5%, and G is 100 generations, then the likelihood ratio of N1 preceding N2 versus N2 preceding N1 (which we can use as null hypothesis with likelihood 1^100=1) is 0.95^100 / 1 = 0.0059. In short, a small degree of relative ease of one evolutionary path versus another may be translated to an enormously strong argument, from a statistical perspective, for that sequence being the easy rather than the even ever so slightly more difficult path. This actually gives us a lot of leverage on the problem.
Human evolutionary stepsThings I think can be ordered in this way are quite a few, in no particular order.
Aquatic Ape stuff. Bipedality. Voluntary breathing control (necessarily preceding speech). Love or tolerance of swimming. Body Hairlessnes. Long head hair. Control of fire. Diet changes. Long distance running. A huge heel. Body temperature regulation by sweating. Ability to swim. Fear of snakes, of sharks, of the various kinds of unknown. Compared with chimps, more acidic stomachs, more ketotic metabolism, less chimp-like raw-fiber digestion, preference for fatter (large) prey. Changes in disgust sensation; Stench, a.k.a. stinky smell sensitivity. Eclectic carnivory; obesity. Tool use. Conversation. Sign systems. Language. Joketelling. Storytelling. Lying. Theory of mind and emotions. Clothing. Multi-sensory integrated spatial awareness. Bodily Decoration. Art. Religion. Moralizing. The Lecture. Admiration. Scapegoating. Mob violence. Male gang formation, predation, war, rape. Multiplication of dominance hierarchies. Gender contrast in anatomy and strength relating to fight and flight contexts. Oversized male larynx. Female lack of sense of direction. Female sexual attraction to funny males. Cryptic fertility of females. Female orgasm. War. Commerce and exchange. Lying. Negotiation. Subpopulation variation in melanin concentration, eye color, height, strength, task endurance. Sexual anatomy & physiology peculiarities: excess size, glans covering, glans skin furrowing, extended duration coitus, semen deposit volume and viscosity and viscosity change over short time periods. Subcutaneous body fat distribution. Literacy. Architecture. Internal storytelling about self as a method for emotional regulation. Various subspeciation changes: more vs less curly hair differences, the Asian epicanthic fold, the northern european adaptations for low-light survival despite metabolic need for light-requiring vitamin D3: translucent skin to make the most of little sunlight and lactose tolerance to get D3 from milk sources.I'm going to concentrate on the part that might make us most human, a sort of conceptual archeology of human language and cognition.
A Conceptual Archeology
Introduction to Conceptual ArcheologyThe following section was written backwards and came out unintentionally and to my surprise. It challenges you to carry out the familiar task of reading in a slightly different way. The repetition of the word "first" may intrigue you into looking back and seeing the relationships, where one idea might break into others. You may find yourself becoming an conceptual archeologist. Have fun, you can do it!
I started at the end, saying "First" (as a linguist would, considering the fundamental concept of the linguistic Sign). Trying, I'd have one view of things and start writing, first X, then Y, then Z. Then I'd realize X didn't start with X but itself includes M, N, O. Then I'd realize M didn't actually start things either but itself started with A, B, C. Then, etc., you see how this goes. So "First" below means: At one point I was thinking in a certain frame and in that frame this thing came first. But you'll notice that "First" isn't actually written first except at the very first, and not even then is it first. So you might think about what you see and can read here, now, as a sort of archeology of the process of writing it, rather than as a digested process of reading it. Maybe that'll be more fun, to follow along with my actual process of thinking, than consuming something fully pre-digested. You of course, get to have your own process of thinking, and maybe putting it in this order will encourage you. So here I write at the first, what's actually the end of my own path of thinking through what's below:
The StrataFirst, we have no idea.
First, we have the possible laws of physics.
First, we have our laws of quantum and relativistic physics.
First, we have the big bang and everything as plasma and quarks flying about.
First, we have hydrogen clouds, coagulating, and nuclear fusion.
First, we have our (first) sun, atomic physics, filling up of the periodic table of elements, and a collapsar supernova event.
First, we have chemistry with heavy elements in a planet around our (second, current) sun.
First, we have a hydrocarbonaceous watery environment with biochemistry and a cycling, available energy source.
First, we have a replication engine and materials.
First, we have accelerated metabolic biochemical pathways to get more done.
First, we have cell walls containing stuff doing stuff replicatably. (See On the Origin of Life, written 14 years ago, evidently in advance of, yet surprisingly consistent with, the latest thinking.)
Functors, Cues, Signals
FunctorsFirst, we have something that acts as a FUNCTOR which is REPEATABLE and USEFUL, which, let's say by definition, separates the doing from what makes the doing go: it initiates (and probably is the first part of) a specialized, available, and useful function, transmitting the information that that function (the rest of it) is to be carried out. So yes at this point the functor itself might also participate in carrying out the function but at least we have some separation between, first, the initiation which represents, by its association with the process, the information that the function is to be carried out, and second the actual carrying out and completion of the action function. So this includes bio-chemical stuff like protein signalling, gene expression controls, RNA machinery triggering and cascades, etc. Viruses already have functors in them, and the stuff inside cells and organelles are no doubt full of them. Already there is lots of information involved, there might be transmission of a trigger within or between cells, there might be some degree of abstraction across distinct but similar and equally effective triggers because the functor is distinct and once activated it cascades to a behavior which is distinct and both are suitable for a variety of related similar contexts therefore there is abstraction.
CuesFirst conceptually is the CUE, which is a functor that is uptakeable, that is to say, it works across a greater separation between the source of relevant information and the distinct behavior which is influenced by it. Being more separate, for example having the source be outside the implementing cell or organ or organism, means that the relationship of trigger to effect is removed from and in a way abstracted from the behavior-implementing process.
Whereas when a functor carries out a biological function through a direct biochemical reaction pathway, we consider there to be information present at the initiation point in the sense that, as viewed from outside, one path is chosen to be taken instead of another, but in the implementing reality it's more that some particular chemical is present in some particular concentration, which makes the changed behavior occur, and only from the outside analyst's perspective acts like information in the abstract, although from the Darwinian perspective it is is the functionality of the response in perpetuating the species which is essential, and as extensively argued above this can also be considered logical or informational rather than merely chemical.
On the other hand in the case of a CUE, there is a greater separation between trigger and outcome along with some form of detection system comprising with specialized detector units, cells, organelles; perhaps there is optionality of the detection event based on organismal priming such as hunger or other prioritized need state; and in this separation we automatically acquire abstraction across stimuli, for if there is any process of detection, it will operate over a variety of stimuli and those might be more or less different yet still inducing the detector to respond, and thereby leading to the same triggered response/behavior/activity carried out by the recipient/responding organism. This pattern, where a variety of stimuli leads to a fixed response, amounts to categorization by the organism.
Given these characteristics of SEPARATION of organism-internal detector from the information source in the external environment, of OPTIONALITY, of ABSTRACTION, and of CATEGORIZATION, we can think of the process as being representational. Thus the existence of cues, or cuing, means the cue-responsive organism implements an abstract categorical representation of external reality.
Said organism might not, it surely does not, contain a motorola microcontroller with bits and actuators and serial ports and compiled code from a high-level language, sure, but what it does can be, and to get the logic right (that is, both, to be survivable evolutionarily and to be understandable by us) it must be understandable as an abstract, categorical, world-representing system. That's what "implements an abstract, categorical representation of external reality" means. Maybe you will wonder what a logic compiler might correspond to in biology.In addition, with optionality we also can consider motivated framing. If the organism might or might not take up a particular cue in any given moment or circumstance based on its own reasons, we can think of it as operating as though it has a directed purpose, a motivated frame, an organized and effective hierarchy of goals, in short an emotional universe, because whether or not there are internal data structures with those labels in the software of the organism, the behavior of the organism will be as if it did. For example, it may chase food when it's hungry but instead will escape predators or toxins even if it's hungry when those are detected, etc.
The logical hierarchy of what we see in its behavior when confronted with different circumstances, that is, what we can infer about its decisionmaking, had better be functional in the Darwinian sense if the organism is to succeed evolutionarily; if it eats even while being eaten then that one might not survive another generation, but if it will postpone eating until it can escape from being eaten, then survival and propagation might be optimized. In this way we need have no idea how the prioritization is done by the organism, but certainly it must do it, for the logic of its universe demands it. Perhaps there is competitive mutual downregulation between alternative behavior-generating processes even within the cell or organelle. The logic of how that competition resolves in different circumstances shows the effective goals and priorities of the organism. That is, goal prioritization and operating within a motivational frame must exist very early in the hierarchy of organisms, so long as multiple outcomes are required for survival and reproduction.
To summarize on cues, they are not just detectable and distinguishable but also understandable within the motivational frame and representational or cognitive capacities of the recipient. A cue doesn't need a sender with an intention but only a recipient responding to a distinction. Cues may be, (N1) first, internal cues as in cell-internal biochemical conditions taken notice of by biochemical pathways which functionally respond thereto, and (N2) second external as in organism-external conditions detected by the organism and used to adjust behavior functionally. (N1 jumps to N2 by more than one step, thus there may be intermediate levels between them so that one could distinguish, for example, cell-external but organism-internal cues, e.g., some cells' characteristics or behavior somehow cuing responses to other cells in an organism, as in organs aiding the homeostasis of the body; or for example, digestive-processing cues that are sort-of-inside, sort-of-outside the organism.)
So to give an example, a male bullfrog being extra big is a cue that female bullfrogs may use to choose it as a mate; there is uptake of information without a sending event; the boy just IS, rather than specifically carrying out a bit of behavior which SAYS. (Labov distinguishes cues from signals in this sense in sociolinguistics, or for example dialect perception.)
SignsFirst (we can stop the conceptual archeology here, this is where it started: was that a fun ride?), we have the SIGN which is repeatable and useful like a functor, which in the event of its optional uptake, abstracts and categorizes a separate external reality, like a cue, and further includes having a FORM which is PRODUCIBLE by senders and detectable and DISTINGUISHABLE by recipients, and also has a MEANING, which again is abstract, useful, understandable within the motivated frame and cognitive capabilities of the recipient. The chicken/egg question between the meaning and the form of the sign is hereby answered: the meaning comes first (in conceptual archeology), for meaning is established in cuing, whereas a producible, distinguishable form is added to cuing to develop, in general, the sign.
Second (finally there is a second, after a first), we have a relatively rich perceptual capacity, so that more signs could potentially exist and be distinguished, though for the moment that capacity may not be used. Here we need fairly detailed perceptual and behavioral systems for this to happen, but that's not too much to ask, since even flies have many eyeballs and bees can dance at many different angles up and down the beehive wall.
Third, the number of signs not just potentially but actually increasing above one so that now there is a vocabulary.
The bee dance might or might not be considered multiple vocabulary items, really it's an analog representation of food source distance and direction, with essentially continuous (down to the resolution level of bees as dancers and as audience). A hard-boiled discretist phonologist insisting on discrete analysis might divide all the horizontal directions from upwind to downwind on left and right into a discrete set, and declare that bees have discrete, phonology-like categorization in their communication system, and since the resolution is limited to what bees in their crowded dance walls can differentiate, the phonologist can never be wrong. On the other hand a continuous-parameter-friendly phonetician might remind everyone we're actually talking about continuous values here, but at a low resolution. You figure it out.
If we're talking about vocal signing, we presuppose
Fourth, we have an analysis of potentially complex signs into parts such as hand position versus hand movement or parts in time a.k.a. sequences, especially the consonant vowel sequence, which is inevitable when the anatomy of the signal source at the glottis is separate from the potential closing points along the vocal tract, such as the lips.
After analysis comes combination so that the different, analyzed, and now combinable, sound or sign forms can be combined as parts together in different sequences, or perhaps in combinations that are non-sequential, such as sign language signs etc.
If a type of organism has evolved to this point then it has the capability of larger, even extremely large vocabularies.
In a vocabulary of signs, logically each inherits the characteristics of being a sign, a cue, and a functor: being useful, uptakeable, repeatable, optional, abstract, expressible, detectable, distinguishable, etc.
Evolution of language above the Sign
After this the evolution of human language is pretty simple, you will start to have multi-word expressions, more than two, and you will have systematic relationships among the word types like for example you will distinguish nouns and verbs and adjectives and then you will use noticeable conventions like order or explicit marking, etc., to indicate some distinguishable relationships like predication or topic/comment structures perhaps in a clear case. I should think that predicate-argument structures would evolve in the order first arguments, then predicates, since a predicate presupposes, only makes sense given the pre-existence of, arguments. Therefore one might consider that Noun evolved before Verb evolved, if nouns are arguments and verbs are predicates, but again, imperative verbs like can have their arguments implicit in the context and thus could be linguistically independent and therefore evolutionarily unordered. Similarly adverbs like Now in [Shoot the Mammoth] Now! can be understood as having the action and the participants all implicit in the living context, so that the evolution of a linguistic modifier need not follow the evolution of its linguistic prerequisite, since that may preexist cognitively. But such caveats don't seem to me to apply here: subordinate clauses could only evolve after clauses; prepositions and postpositions only after the phrasal units that they include as required constituents.
Next, something might not quite fit with the usual interpretive schema for most cases, but with that schema or form or structural relationship regularly used and established its usual circumstances it can still usefully be re-used, the form imposed on the ill-fitting case, and users may go ahead with it, and now the linguist will extract required structures that fail to have consistent semantic interpretation and that means phrasal syntax.
Finally you get Chomskyan Merge which gives recursion. You can't recurse on a category unless you already have that category. Thus we have ordered quite a lot of human language evolution into a sequence of developments, at each point there being a coherent useful and structured system benefiting its users by some higher degree of getting on the same page, social coordination, and thus better survival and reproduction for those capable of it. And by now you have human language, with vocabulary, syntax, and semantics.
Next, we will touch on pragmatics, and finally follow along again with the evolution of phonetics, which is very cool.
Pragmatics before the SignPragmatics of course has to exist once sign interpretation is optional or you might say voluntary, and since optionality is present in cuing systems, it follows that pragmatics had to exist even during a cuing-only stage of species development. Does the cue imply cooperation with a con-specific? A receiver might calculate the reliability or unreliability of the sender's cue, might infer alternative conclusions from those normally inferred from the cue, might have an effective representation of 'self' and 'non-self' or even of 'other' and a world of other impinging abstractions with which to effectively reason about the cue and its meaning for the receiver. These are pragmatic possibilities before the linguistic sign ever arises.
Three of the four Gricean implicature rules in the field of pragmatics seem to fail to apply within the universe of cues. Consider (1) quantity, (2) quality, (3) relation, and (4) manner.
(1) FAIL: Does the cue, if it includes a QUANTITY, imply that the maximum, or perhaps the minimum, quantity involved is that indicated by the cue, or just that the cued quantity is within some inferrable but unknown range? It seems that the quantity cued is exactly the actual quantity in the cue itself, so that implicature toward set maxima or minima are neither relevant nor reasonably, customarily, inferrable.
(2) SUCCEED: Does the cue imply the QUALITY of truthfulness? We do see species that try to fool each other, but cues wouldn't have value if there weren't useful, i.e., sufficiently true, information involved.
(3) FAIL: Does the cue have proper RELATION to the goals of the cue-detecting organism, or is it rather an irrelevance or distraction? Since a goal-directed organism may need to explore in the weeds until it finds what it needs to move forward, this Gricean issue of relatedness, relevance, or pertinence, seems to be in the eye of the beholder, and not packaged therefor by any cue-er. The maxim of relation applies only at or above the level of the sign, it appears.
(4) FAIL: Does the cue provide the needed information in a clear, brief, orderly, evident, and unambiguous way? Again this is in the eye of the beholder. It is the job of the perceiver to perceive what is useful in a world of noise, whether that is an easy job or a hard one. So the maxim of manner again seems to be useless in reasoning at the cue level, therefore applicable only at or above the level of the sign.
So much for pragmatics.
Let's have a evolutionary run-through with an eye on phonology. In the above conceptual archeology of increasingly human language like systems, there never appeared a moment in which the brain had suddenly evolved to incorporate a new silicon or other discrete-logic-based module comprised of a digital, binary, finite state, representational machine for the phonological system with established bits for labiality and velarity and nasality etc. Such models are scientific abstractions, however pretty to phonologists. No, on the contrary there has long been, before combinatory signing, a given, pre-existing, anatomical and functional production-and-perception system capable of making a variety of detectable, distinguishable, reliably producible sounds, repurposed from the mammalian aerodigestive, breathing/eating/puking orifice and passageway which obviously was already multi-purpose.
Evolution of Phonology
Let's take one phonological feature, the glottal state bit known in Chomsky's Sound Pattern of English as [Voice]. Where did the glottis come from, indeed, this sound source separate and deep in the vocal tract? It goes far back on the evolutionary tree to land animals with lungs. Does it go back to gilled fishes? Once you have lungs (or gills) and you double-use your food orifice for respiration, you will also need a larynx. So in the reptile aero-digestive tract, the larynx is a simple breathway tee-valve to keep the food out of the lungs, its structure being the larynx itself, and its closing elements being the ligaments of the glottis.
This weird buzzy widget develops and ramifies miraculously across species. Who knew?! In a variety of mammals, muscles can jam the larynx right up inside behind the nose, with the glottis open, so that nasal sniffing, smelling, and breathing can all go on continuously while the mouth is being used simultaneously to suckle, swill, or strangle, to catch, carry, or chew. It is freaky to imagine.
Kids, don't try these at home! Our human larynx of course won't allow it. We have to either breathe or swallow, not both at once, not safely. Because our larynx doesn't plug up inside the nasopharynx in humans, instead it hangs low, between a relatively lengthened oral cavity and a shortened esophagus, and the epiglottis and glottis redundantly cover and protect the airway tee branch to the trachea and lungs from drippings or food. Worse, the larynx is grossly enlarged, and even more lowly hung in the male of the species, but in either sex making a supra-glottal resonant chamber even better for loud vibrating sounds. We call them vowels. But it was long before humans that the glottal ligaments became evolutionarily useful in noisemaking, loud noisemaking, on behalf of their owners. For dogs bark, prairie dogs yip, rats squeal, all these providing noisemaking power antecedent to the ramifications of phonology. Not that the lowered and larger larynx didn't further develop with the same evolutionary benefit of more resonating power.
Yes, given that the inherited system could make some sounds is a good enough starting point.
But before that had to evolve the aero-digestive oral tract, with a digestive passageway, passageway for air, passageway separator (larynx etc.), tongue, teeth, and lips, along with sub-oral components like the digestive tract itself, or the lungs and the diaphragm to push on them. First a digestion capability to transform raw environmental findings to incorporateable nutrition. (First there was exterior digestion: there exists some beast that digests prey outside its body, perhaps an octopus?). Then an enclosure for digestion, which being an enclosure had to comprise a cavity (to hold the work) and a mouth for controlled transit (letting in (eating), letting out (excreting), keeping in (preventing escape of the doomed), keeping out(preventing ingress of non-foods).
First, we might could benefit from a passageway through, to speed up and specialize processing steps for food, since a single sac with regurgitating of waste like jellyfishes means all the steps have to be available in the same space and they may interfere in time, space, or process, making the digestive process inefficient or weak. To the rescue, perhaps a single mutation could double the enclosure count from 1 to 2, with a connection between, with an initial efficiency advantage from the fact that now stuff can be ejected from one end while being ingested at the other, being faster than intermittent operation of swallow-then-spit. The passageway and other opening being advantageous from the start, specialization of digestive processing at different points along the passageway might evolve to make the system even more efficient, able to get more nutrition from worse food, thus expanding the reachable universe for life. And so we discover.
Among Tongue, Lips, Teeth, what came first? Was it the tongue first to position and move food inward? Then lips second to contain food and keep it from backing out, which is logically second because you don't need to contain what you can't position and draw inward. Do jellyfish have lips? Teeth third to improve the food processing part. Sharp hard bits have their uses.
I am a bit confused about lungs and voluntary breathing through the aero-digestive tract. Perhaps evolution was, as we might say in personifying its actions, trying a wild variety of options as tetrapods moved from water to land. I suppose fish with gills already have mouth-to-gill ventilation of the organs; and I thought some fish have external gills, well how do those move down into the abdomen? Evidently the creatures might be more able to be out of water by having air-oxygen incorporating tissues on their outside. Then cover these air-gills for keeping them suitably wet and uninjured so that it can bash about on land survivably. Then provide active yet non-distracting ventilation for them, for even more safety, perhaps back out the mouth rather than through gill outlets in the side of the neck. Not sure how that proceeded. It seems a pretty unlikely, even a sharp mutation to move an organ from the outside of the neck to the inside of the ribs, but once there it can more gradually evolve a trachea, brachii, and lung sac alveoli, but by now you essentially have evolved lungs and you can climb onto the land.
Could the shutoff valve larynx have evolved after the aerodigestive tee shaped tract? Certainly it would be useful in such a context. To solve the chicken-egg problem, perhaps there were two passageways competing for the narrow neck transit room, say we'll call one opening a "nose" when it's the air passageway, and we'll call it a "mouth" when it goes to the food passageway, and then there needed to be some stiff, cartilaginous, membraneous structure just to keep their tubes apart and allow simultaneous use despite both sharing the tight spaces of a neck. Once it became a single organ, with two inputs and two outputs, it could simplify to become a tee valve, adding some evolutionary value at this step also. Because it is advantageous for air and food to be able to share a passageway, and then the fact we have both mouth and nose today would be an evolutionary holdover from when the mouth couldn't be used for air. Somewhat persuasive, but I don't know any physical archeology that bears on this. It would seem to be early, and maybe we're talking about when we were still fish.
In conclusion, we are evolution's slaves: its leash is tight, and it kills across whole species, unblinking, and controls us fiercely, down to the shapes of our cartilages with a cold and analytical heart. Every feature in every species that comes down into posterity as a feature of the species only took hold after innumerable mutations that failed to survivably coexist in the species, after that one proved survivable still had to multiply against valid alternatives, still after being endemic in the species had to show an advantage, and in showing an advantage such that the whole species now has it, something like an entire species worth of childless uncles and early-killed lovers had to be consigned to death, perhaps eaten, but certainly decomposed and recycled without living trace, for the spread of that feature to be ensured and for the carriers of its alternates to no longer exist in the species. Every species feature has cost a species-wide genocide of early death, or organismally unfulfilled life without offspring, for that feature to now characterize the species. How many thousands and millions of features we see in the amazing natural library of the species of the world is a testament to accumulated death and loss in an unimaginably long and bloody past. We ourselves are not even blips in the flow, in the cascading wash of life. Let us be humble as befits our station.
Was that as fun for you as it was for me? Hoo!