…- A Sign of Unintelligent Design
In a recent thread here at TSZ William J. Murray brought up the subject of semiosis, to which Allan Miller responded:
Allan Miller: Nice to see WJM has absorbed UB’s ‘semiotic’ nomenclature though. Published nowhere, of course…
It never ceases to amaze me how critics of ID will say anything just be contrary regardless of whether what is said is true or even coherent.
Allen then goes on to describe how he thinks a semiotic system could arise HERE.
So perhaps Allan can be the very first to publish in this new and exciting field of semiosis. More likely though, is that his initial claim was just false. (Probability near 1, IMO.)
How does Allan know what elements need to be present in order to establish the existence of a semiotic system?
There are in fact some interesting comments over in the thread that initiated this OP that I may in time copy over here.
Any claim that Upright BiPed created the terms semiosis, semiotics, biosemiosis or biosemiotics is simply ludicrous.
The claim that WJM has absorbed UB’s ‘semiotic’ nomenclature is likewise ludicrous.
The claim that there’s been no publishing in semiosis, semiotics, biosemiosis or biosemiotics is simply ludicrous.
People who make such claims should not be taken seriously.
[title split by Lizzie]
Let’s start here:
Essential Readings in Biosemiotics: Anthology and Commentary
Note that’s Vol 3 in a series. By a major publisher.
cf: http://link.springer.com/search?query=Biosemiotics
Simply absurd [ignorant] to assert that this is a field which is lacking in published articles or books.
Meanwhile, it is not at all difficult to multiply examples, a single one of which serves to falsify the claim.
Introduction to Biosemiotics: The New Biological Synthesis
See also:
http://en.wikipedia.org/wiki/Biosemiotics
Elizabeth:
WJM:
Elizabeth
WHAT is not a semiotic system?
Elizabeth
Allan says you’re wrong. WJM says you’re wrong.
The questions you were asked were:
Where else in nature can we find a semiotic system? Your answer was nowhere else. So you agree with WJM and disagree with Allan.You cannot both be right.
And:
Tell us in principle how nature can generate a semiotic system. Your answer was … nature did generate a semiotic system … therefore in principle it can.
Are. You. Serious?
It’s crystal clear to me that Allan Miller was referring to the UD commenter, Upright Biped, and his “semiotic argument” when he wrote:
which has been discussed here at some length, sometimes involving Upright Biped. The suggestion was made by many that UB should publish his “argument” if he thought it had merit. This is yet to happen.
That there is a respectable literature in semiotics and biosemiotics is not and has never been an issue. This OP is a waste of pixels.
IOW, you’re bitching that Upright BiPed did not publish materials already published. And you accuse me of wasting pixels.
You’re ignorant of it and therefore it’s not an issue. Yes, I got that. Perhaps you should read up.
Or, given your intimate familiarity with the literature, you can of course refute Upright BiPed’s position from it.
Not. Going. To. Happen.
Ever.
Tell us, Alan, from the literature, what is a semiotic system?
Is the cell a semiotic system?
Depends how you are defining “semiotic”. In the accepted sense of the word, I would say no. Is a cell a living system? Depends on how you define “living”. I would say it is irrelevent to the study of cells. My interest in a cell would be “how does that work?” Answering “it’s semiotic!” or “it’s living!” doesn’t add much in the way of understanding.
Word games can be fun but they don’t much help in scientific endeavour where clarity of thought trumps obfuscation.
Yeah. They are just semiotics.
Semiotics is the study of signs and symbols (the word being coined by the philosopher John Locke around 1690). Biosemiotics is a cross-disciplinary (which I’m all for) approach to the study of signs in biological situations. UB’s argument is word salad.
And your claim that Upright BiPed’s argument is word salad is due to your ignorance of the literature.
I not regard the molecular system by which cells produce proteins and enzymes a “semiotic” system by the usual definition of “semiotic”. The pairing between codon and amino acid works the way it does because of the existence of the specific set of tRNA molecules that do the mapping. Sure you could use a different mapping, but in that case you’d need to fabricate a specific and different set of tRNA molecules to achieve it. In other words, the codon-amino acid molecule mapping is determined by the specific set range of tRNA molecules available. And that set is itself a result of the physical form of the DNA.
You might argue that the system is too complex to have evolved, but not that it is too semiotic to have evolved.
I do think that biology is capable of producing semiotic systems, however – because of course I think it produced us, as well other organisms that appear to be capable of communicating by means of symbols.
But when we use symbolic systems to communicate with each other we really can be entirely arbitrary in our choice. If I want to change the mapping between, say the word “rabbit” and the animal to which it usually refers, so that the word “parrot” will now indicate to you that I mean what I used to mean by “rabbit”, I do not need to do anything other than tell you, and you don’t need to do anything other than remember. And it won’t matter, thenceforth whether I tell you about the animals-formerly-known-as-rabbits by speaking to you and using the word “parrot”, or writing to you and using the letters p a r r o t, or even P A R R O T or morse code, or semaphore – you will still know that I am referring to a small furry mammal with long ears.
I do not have to design a special molecule to deliver the word parrot to you that will cause you to think of rabbits, and not parrots, when it arrives. But this is what we/a designer would have to do to change the mapping of GUC from valine to glutamine – we’d have to rejig an RNA coding stretch of DNA to produce a new tRNA molecule with a GUC bond at one end and a glutamine bond at the other, and instead of the old GUC-valine tRNA molecule.
That’s because the mapping is governed by chemistry, not by an arbitrary agreement made between two symbol-using human beings.
LoL. Relativism comes home to roost. Thank you.
There is no link between UB’s word salad argument and the discipline of Semiotics. UB just throws in some words to his mix.
Elizabeth,
Great Crested Grebes and Badgers!
Mung: Is the cell a semiotic system?
How I define the term is irrelevant. There exists an established literature on the subject.
What, according to you, is the “accepted” sense of the word?
Allan Miller claimed to have provided a path to a semiotic system. Is he wrong?
That’s true! 🙂
Real life intervenes. Back later.
Not relativism, Mung. Semiotics.
You have empirical evidence to substantiate your claim but you think it’s for the common good that you hide it from us. We should take it on faith.
Think on that, Elizabeth.
I have.
1. What is the usual definition of semiotic?
2. Why is Allan Miller wrong?
So semiotic systems are characterized by mappings, e.g., codes. The mappings are arbitrary. This refutes Upright BiPed how?
The usual definition of semiotic, is that it is to do with the communication between agents by means of signs or symbols where signs and symbols are patterns that carry meaning. A sign (also sometimes called an “index”) is sometimes differentiated from a symbol in being some kind of “pointer” – having some kind of inherent meaning that can be readily understood without agreement, whereas a symbol requires agreement between the communicating parties.
So if an alien arrived on earth, and pointed an appendage at a tree, even a human might understand that it was trying to indicated to us that it was interested in, or wanted to direct our attention to, the tree.
However, if it were to say “qwakndfff paodm qerqr” we’d be utterly mystified because we do not know what those symbols mean.
And so the DNA-mRNA-tRNA-amino-acid mapping is not a semiotic system, in that sense. There are no signs and no symbols. There is no agent communicating meaning to some other agent.
What there is is a mechanical system, which, given a certain physical starting point, results in a certain physical end-point. The only way to change the mapping of codon to amino acid would be to physically design a different tRNA set that would do it, and physically alter the DNA so that a different tRNA set was produced.
If you, or anyone else, wants to extend the definition of semiotics to include systems in which the physical characteristics of the communication medium determine the mapping, that’s fine. Musical boxes pr piano rolls now become semiotic systems under that new definition.
But in the past, musical boxes and piano rolls were used as the paradigm cases of a system in which there was a mapping, but not a semiotic one. A piano roll could be used as a symbolic symbol by a skilled musician – it would be possible to regard the holes as a kind of notation, in which case black dots in place of the holes would do just as well to tell the player which notes to play. But to the player-piano itself, it is not. Those holes have to be holes or the system won’t work, because the mapping is mechanical for the piano, not semiotic.
Well, it doesn’t “refute” Upright Biped, because I don’t think Upright Biped even makes a coherent case. But it reveals the incoherence, in my view.
The DNA-mRNA-tRNA-amino acid mapping is physical. To change the mapping you have to change the physical properties of the parts. In a semiotic system you don’t – both ends of the communication line simply agree to use a different symbol, which can have any physical properties they choose.
Could you point to where anyone made that claim?
Well he may have come to the conclusion completely independently, but I suspect that time spent on UD is the principal reason why he would choose to term the DNA-RNA-protein system ‘semiotic’. This is not a common usage – it refers more typically to signalling. The translation system is not doing anything like that, and I’m not aware of any publications that follow UB’s usage. [eta – I’ve done a quick Google, and now I am! Nonetheless, as a ‘theory of ID’, and a challenge to evolution, I’d be pretty sure that WJM would not be saying ‘semiotic’ wrt the genetic code but for UB].
Absolutely. Don’t dignify their stupidity with its own OP, whatever you do! 😉
Just to be clear, I was happy to agree with WJM’s definition for the sake of simplicity. If he wants to call translation semiotic, and then demands how such a system can arise, I am more than happy to oblige with a sketch.
It’s not ‘really’ semiotic, but since I understood what he meant, there was no point in pursuing definology down the rabbit hole. Else I’d be accused of avoiding the issue with pointless arguments about words.
If we want to talk about systems more within the realm of biosemiosis – signalling for example – we can talk about that instead. But he wanted to know how protein translation could evolve, as I read it.
I think it’s irrelevant anyway. No evolutionist disputes that evolution is capable of producing “semiosis” otherwise they wouldn’t claim that it produced human beings.
Where you want to call what happens during protein translation “semiosis” (and I wouldn’t ) is neither here nor there.
I still come back to my point that if a Designer was involved in tinkering around with nucleotides, then that Designer was exerting force on molecules that opposed the forces we know about.
Or, alternatively, selecting from the range of all possible universes one in which the molecules just happened to produce people. but in that case we wouldn’t observe, from within in it, anything non-natural about the process. But most ID proponents seem to have in mind a Designer that reaches in, from time to time, and moves stuff into otherwise improbable configurations.
Skimming the Biosemiotic journals, I have yet to find much of substance. It is presented as an “entirely new conception of biology”. “Its aim is to strike at the very heart of the life sciences”. Hmmmmm.
Signalling in biology is really interesting. I don’t see that anything is gained by dressing it up in Saussurean language. Just confuses the issue, and does a disservice to Saussure.
Elizabeth,
Sure. Not all that new, though. One can extend the concept of ‘signal’ past usefulness. Is the concentration of an enzyme product or substrate a signal? Kinda. It can induce conformational changes that slow down or speed up production or consumption. But …
Mung,
The Semiotic Theory of ID has been published? Where?
Well, I think that it’s a potentially useful concept when considering systems, and organisms certainly be considered as system in which a heck of a lot of information transfer goes on.
I just wouldn’t call it “semiotic” because it is all accomplished by means of direct chemical/ionic reactions.
Ultimately, UB conceded that it does not follow from his characterization of protein synthesis as “semiotic” that it must have been authored by an “intelligence,” or that it could not have arisen naturally. See comment 1001, here.
His fallback position has been to assert this mistaken argument:
My response
UB:
RB:
UB never surmounted this difficulty, in my opinion.
One important point is that genotypes that reliably produce phenotypes will tend to evolve more rapidly than genotypes that do not. In other words, features that make the fitness landscape smooth (similar genotypes => similar phenotypes) will tend to be present in populations capable of adaptive evolution, and thus, at population level, will tend to be selected.
Therefore once you have a tRNA system with minimally reliable mapping, it will tend to evolve to become more reliable. It will also tend to select many-to-one redundancies (many codons for one amino acid) over one-to-many redundancies (one codon for many amino acids).
And indeed, that’s what we observe: several many-to-one redundancies, but no one-to-many redundancies.
Contrast with human language where many signifiers can have the same referent (pigs, hogs, swine; porcines), and a single signifier can have several referents (pigs: porcines; police; greedy people; blocks of smelted metal).
One of the interesting things about Darwin’s fundamental insight is that can nest at many levels – evolvability itself can evolve, which may well be the clue to the early evolution of replication itself. Not only features that promote individual breeding success will become more prevalent, but features that promote adaptive evolution populations will also become more prevalent.
You have empirical evidence to substantiate your claim but you think it’s for the common good that you hide it from us. We should take it on faith.
Governed is not the same as originated. The origination is what is in question. The fact a mapping exists implies chemistry cannot originate the mapping anymore than hardware can originate software. The hardware can only enforce a mapping, it cannot originate it. If chemistry originated the mapping, the notion of mapping would be superfluous or meaningless.
Though the mapping is presently governed by chemistry, experiments and observations imply it is not originated by chemistry — as illustrated by Humpty Dumpty.
The mapping can be implemented differently as evidence by alternative genetic codes and and alternative start codons.
For example the leucine codons “CTG/CUG” and “TTG/UUG” can code for methionine.
http://en.wikipedia.org/wiki/Start_codon#Alternative_start_codons
http://www.ncbi.nlm.nih.gov/Taxonomy/Utils/wprintgc.cgi
Trying to distinguish between “hardware” and “softwared” at the molecular level (or even, in the context of brains, the neural level) is at best useless and at worst misleading.
There is no intrinsic reason, as I explained, why an initially unreliable but minimally functional mapping would not evolve into a reliable one, given that a reliable mapping between genotype and phenotype aids adaptive evolution.
So sure, chemistry can generate the mapping – chemistry generates the genetic changes that fiddle with the mapping, and better mappings will be replicated more often, at population level.
Interestingly, there can, at least in theory, be such a thing as too good a genotype-phenotype mapping; a bit of “slop” between genotype and phenotype can help ensure richer range of genotypes in a population, and thus render it more robust to environmental change.
And Humpty Dumpty doesn’t help you at all. As Mike Elzinga repeatedly points out, at molecular level, Humpty’s shell isn’t even brittle.
stcordova,
Experiments and observations imply … ?
Suppose I liquidise a radish. It does not reassemble into a radish. What does that tell us about the origins of radishes? Nothing, I would suggest. Perhaps I broaden my analysis to include the rest of the vegetable kindgom. Again, probably no spontaneous reassembly. OK, animals now. Fungi, bacteria … no, nothing. So the origin of life is perhaps not due to the spontaneous reassembly of dead things, then.
What is it about ‘Humpty Dumpty’ that makes you refer to it in every post you write? The origin of life (which I would take to mean the origin of replication) took place on a messy earth, in some environment or other combining the physical, the chemical and the dynamical. It would be rather surprising if it just happened in a test tube.
If Sal’s point had merit, crystals would be impossible.
stcordova,
In almost all variant codes, the variant codon is, in at least one of the variants, a STOP. This tells us something very significant about the amendability of the code. It tends not to admit of replacement of one acid by another (nowadays) but the extension of the tail of a protein is less destructive. It is probable that an acid cannot become a STOP either – this would lead to serious fragmentation of peptides. So the most likely move is STOP -> Assigned.
The implication of all this is that a code full of STOPs (perhaps with only one or two ‘actual’ codons initially) can readily become a code full of assignments by evolution, replacing STOPs by assigned amino acids.
Using the word “minimal” is misleading. More than one OOL researcher has had the uneasy intuition there might be no such thing “minimal” (as in trivial) if we he have a DNA-RNA-protein replication cycle. It would have to be extravagant to begin with because of the chemicals involved.
Hardware is the laws of physics and chemistry, software is a configuration among the possible legal configurations under those laws. The analogy is quite accurate.
The problem of OOL is the origination of extravagant configurations (software) that are interdependently complex to achieve things like replication and metabolism.
Correction
Allan
No, replicators easily arise if the replicator naturally occurs under a wide variety of circumstances. i.e. the image of a tree is easily “replicated” by reflection off water.
The OOL problem is the origin of theoretically improbable replicators such as those implemented by life. OOLers keep redefining away the real problem!
One such theoretically improbable replicators could involve:
1. reading off a memory tape, with starts and stops
2. the read and write mechanisms are also described by the memory tape (at least in substantive part)
3. the material the tape is made of can physically be configured many ways, but only a few of those ways would lead to replication
The OOL problem is that this sort of replicator (von Neumann Constructor approximation) is implemented by life.
The Humpty Dumpty illustration is just an illustration to remind the readers that even with all the chemicals present in the right proportions, such an intricate extravagant replicator will not spontaneously arise if even a few key components malfunction. That highlights the high probability of evolving non-functioning states vs. the low probability of functioning states from a dead state. The dead state of Humpty Dumpty evolves to other dead states and even farther from life because of biological and chemical decay mechanisms starting with decay of life-critical homeostasis.
Well, a successful OOL theory would have to start with a self-replicator simple enough to form spontaneously from non-self-replicators. If no such minimal thing is possible, then OOL doesn’t work. But not knowing what that self-replicator might have been is not the same as saying that a simple spontaneously emerging self-replicator isn’t possible. Crystals, for instance are form of self-replicator, so are bubbles, and some molecules self-replicate in the presence of catalysts. The most promising scenario on the table right now, I think, is a combination of bubbles and self-catalysing polymers of some sort.
OK, but in that case I think it’s a poor analogy. Use it if you want, but beware its limitations. Cells don’t run software.
But how can you reprogram cells if they don’t run software. 🙂
http://www.mccormick.northwestern.edu/magazine/spring-2014/synthetic-biology.html
Just point out, I’m not idiosyncratic in referring to life as having software, it’s the OOLers and evolutionary biologists who dislike the appearance or insinuation of “program”, because program looks a little too much like purpose.
stcordova,
Afraid I see that as a pointless ‘gotcha’. You know what I meant.
How so? I referred to the OoL as being the origin of a ‘true’ replicator – a system that can replicate itself with sufficient fidelity that the copy is also a replicator (even if inexact). I’ve never wavered from that.
I think the properties of nucleic acids are fundamental. They have the unique property of possessing a kind of bootstrapping – of ‘self-specification’. It is not incidental that the 4 bases in RNA/DNA form 2 sets of complementary pairs, out of all the bases that could have been used. I think this is in itself fundamental. The first requirement, even before replication, is stabilisation, and dsxNA provides that. It is quite remarkable how readily complementary xNA sequences find each other. It happens with RNA probes, in more than one way in PCR, in hybridisation analyses, in mRNA silencing by complementation, in strand annealing repair pathways … antiparallel complementary strands have high affinity, entirely due to physics. So I think, in the first instance, you can forget ‘algorithmic’ replication. Single RNA strands tend to cyclise – the 3′ and 5′ ends meet up, to form a loop (see also bacterial genomes). But a random mixture of such short cyclic RNAs will tend to pair up in complementary manner. This increases the half-life of the strands significantly over single strands, because hydrolysis is rendered less likely, and such hydrolysis as occurs can re-ligate because the lattice is pinned. More stable RNA crystallises out of the mess by virtue of this chemical ‘selection’. And for free, you get homochirality, since homochiral pairings of either ribose isomer will be more stable than mixtures.
Now, getting this towards replication or production of ribozymes (close to being the same thing) is of course no easy task. I don’t suppose you’re really interested in how the OoL might occur naturally. But such scenarios render it substantially less implausible than your ‘algorithmic’ cartoon.
So we need a non-equilibrium, dynamic state – a feed-in of localised energy. Got that.
stcordova,
More that it leads people into over-extension of analogy. Systems are what they are, not what they are a bit like on some level.
That’s not the reason, Sal, It’s because it is a potentially misleading analogy. The “hardware” and the “software” are not separable in living things. You don’t have a dead body, and then install software on it to bring to life. In organisms the two are inseparable – the very structure and mechanisms that an organisms consists of are also the “software” that makes it grow, repair itself, nourish itself and do whatever it takes to breed copies of itself.
None of that has any analog in the “computer hardware running software” analog.
What I think many ID proponents miss is that ID critics are not arguing that organisms aren’t awesomely complex bits of the universe. Of course they are! That’s why biology is such a fascinating subject.
Organisms are more impressive than computers, and better “designed” – by dint of billions of years of optimisation – to thrive in the niches they occupy.
Billions of years of damage including catastrophic extinctions. You think time improves the chances of construction, it actually improves the chances of destruction. I pointed to specific chemical reactions that have been proven in the lab.
Darwinism and chemistry don’t mix, especially in a pre-biotic Earth where even Darwinist like Dobshanzky said selection doesn’t apply.
You can’t optimize by Darwinian selection populations that are dead.