Private perceptions and public images

[ilkka_nissila]

Ups, I missed the last two posts. I would like to emphasize that in neurons, the membrane potential fluctuates up and down depending on input, and this is clearly an analog signal. It is trhresholded, yes, and this determines if the impulse is generated, but in a computer the input signals are binary and thresholding a binary input (represented by an analog signal which is one of two voltages) is a quite different thing from thresholding a fluctuating analog signal. If you have two excitatory inputs to a neuron, do you really believe that it just adds them up? Although an ion channel is triggered by a certain threshold voltage level, and it remains open for a precise time, the number of ions that transfer across the membrane during the opening time is not the same every time. And since the membrane potential fluctuates the triggering of a subsequent impulse is definitely determined by an analog process (although you might consider it discrete because ions are discrete units). The number of ions is so large and there is so much variability that the outcome is just not determined by the inputs alone.

[pico]

<i>Unlike the computer, the brain is very flexible. It changes itself in response to external inputs. Its calculations are erroneous.</i><p>

 

There are many computer applications which can adjust themselves, and programs that rewrite themselves. If you want a glimpse, look at Danny Hillis' <u>Patterns on the Stone</u> in which he describe a program that writes programs. It is given a model of the expected outcome, for example a sorted list, and input. It rewrites the code until the outcome is made true for any input. Funny thing, but it even came up with an inscrutible novel technique.

[ilkka_nissila]

Pico, my point is that the brain is changing all the time and this happens at the basic building block level. How many microprocessors change the way they work? Yes, software can be made which changes itself, for sure. But this is not the way computers normally operate.

[pico]

You are speaking of today. Tomorrow will be different.

[andy_laycock]

Ilkka,

 

You are missing my point entirely which is that the eye is more like a digital camera than film. Not only is it regenerative but also the transfer of the information from the eye to the brain is by action potentials, which are discrete and repeatable. Action potentials are more 'digital' in nature than analog. I also said there was a great deal of modulation which is what you are mainly talking about. But that modulation is concerned with whether an action potential is going to be generated or not and at what frequency. There is also a great deal of modulation in digital electronics.

 

As for Pico's discussion on quantum mechanics I'm not sure why that's even being brought up, especially in such large cellular structures as microtubles. To me it just seems like a red herring.

[pico]

<i>You are missing my point entirely which is that the eye is more like a digital camera than film. </i><p>

"than film"<p>

Talk about a strawman. The eye is more like a digital camera than a loaf of bread, too!<p>

But in reality there are more differences than similarities. The eye is not a camera. The eye is not like a digital sensor. The eye is an eye and a sensor is a sensor.

[Edwin Barkdoll]

This is probably one of those discussions best avoided...

 

I would argue that there are interesting similarities and differences between a biological sensor like the retina and electronic and chemical sensor arrays like CCD's and film, but that arguments about which of the latter two are most like the former are probably unresolvable and beside the point.

 

That said, consider the first stage of each:

Retina involves a destructive, photochemical process (photoisomerization of rhodopsin). This is not "spontaneously" regenerative at the molecular level until the chromophore is chemically regenerated.

 

Film involves a destructive, photochemical process. This is never "regenerated" until another sheet of film is put in.

 

CCD involves non-destructive production of charge by photon absorption. Regeneration is unnecessary.

 

Here I would say the nod goes to film - destructive and photochemical.

 

Next, first stage sensor "output":

 

Retinal photoreceptor (vertebrates here, inverts can be different) output is strictly analog (neglecting quantized output at low light levels due to guantal nature of light).

 

Film is "digital" if I can use the term that way. That is, film silver halide crystals are either converted to metallic silver or not depending upon light absorption.

 

CCD individual sensor output is analog. Subsequent analog to digital processing produced the digital signal.

 

CCD (e.g. digital camera) gets the nod here - analog sensor output.

 

One could argue that the retinal cells prior to the ganglion cell stage, which outputs digital signals to the brain, operate *in part* as A/D converters.

 

Really tho' any mention of film versus digital outputs of retinas and CCD's is inappropriate - film is a sensor array with no inherent processing beyond the first stage unlike retinas and digital cameras.

 

Finally areguments whether neurobiological systems such as the brain, eye etc are either digital or analog miss the point. Clearly analog components (e.g early sensory processsing) and analog processing (for example the massive dentritic arborizations of cerebellar Purkinje cells are primarily analog signal processors) are essential as are also the digital signals of the nervous system, the action potentials.

[ilkka_nissila]

Andy, where is "modulation" in digital electronics? I don't get it.

[andy_laycock]

"...analog processing (for example the massive dentritic arborizations of cerebellar Purkinje cells are primarily analog signal processors) are essential as are also the digital signals of the nervous system, the action potentials"

 

= modulation of digital signals

[Edwin Barkdoll]

"= modulation of digital signals"

 

Not really a helpful or even meaningful comment, rather simply a redefinition of terms.

 

According to your logic if, in this case, analog processing is simply modulation of digital signals, then just as helpful a redefinition is digital processing = modulation of analog signals.

 

Since all first stage sensory output is analog as is also most input to axon hillocks and other sites of action potential initiation (negelcting things like nodes of ranvier which are not relevant) I would argue that if either, digital processing = modulation of analog signals. Really tho', neither restatement adds anything.

[ilkka_nissila]

Andy, you didn't answer my question. Earlier you wrote "Digital electronics also have a great deal of analog modulation as well" (comparing electronics and the brain).

 

Where is this analog modulation in digital electronics?

[harry_flaxman]

Very curious indeed. Analogue is analogue...dgital is digital. There are no two ways about it. The brain operates on chemical transmissions...not digital transmissions. The camera operates on mechanics alone.

[al_kaplan1]

This has sure strayed from Billy's question at the start of the thread, which han nothing to do with whether or not the brain is analog or digital. When we look at a scene we want to make a photograph that evokes the same feeling as we're experiencing (private perception) as well as just recording it (public image). Is this possible?

 

Some photographers are better at evoking feelings than others, but that just begs the question: are they the SAME feelings? I think that it might not matter all that much as long as the viewer gets pleasure from the photograph.

[andy_laycock]

Ilka,

 

anything circuit using capacitors or inductors could be analog modulation.

 

Yes the chemical transmission in the brain is chemical in nature and does not use electrons. However the AP, which is the transmission that I'm referring to is 'digital' in nature. Either it happens or it doesn't. There's no in between. It's the major tranmission source for information along axons. You can delve as deeply as you like into all of the biological systems involved in signalling (such as second messenger systems etc.) but you won't alter the fact of the AP. That's the last I'm going to comment on this. I stand by what I say and nothing is going to sway me, so live with it :P

[ilkka_nissila]

<i>anything circuit using capacitors or inductors could be analog modulation. </i>

<p>

Digital electronics doesn't work that way. The presence of inductors or capacitors is irrelevant. Digital electronics is implemented in such a way that the signal level is always far away from the threshold between 0 and 1. When a signal switches from 0 to 1 or 1 to 0, it happens during a part of the clock cycle which is such that the intermediate values of the signal voltage are not able to affect output. There is no analog modulation unless you're talking about measurement systems which have analog-to-digital converters. But that's at the threshold of analog and digital electronics.

<p><i>

However the AP, which is the transmission that I'm referring to is 'digital' in nature. Either it happens or it doesn't. There's no in between. It's the major tranmission source for information along axons. </i>

<p>

I don't dispute that neuronal impulses are digital in nature. But the decision whether to fire is determined by complicated biological processes which are NOT digital. This is the key difference: digital signals are only used in the brain for signal transmission. Computations are analog. In digital electronics, computations are digital. They're carried out by logical operations with 0s and 1s.

<p>

<i>You can delve as deeply as you like into all of the biological systems involved in signalling (such as second messenger systems etc.) but you won't alter the fact of the AP. That's the last I'm going to comment on this. I stand by what I say and nothing is going to sway me, so live with it :P

</i>

<p>

I recommend that you pick up a book on computer architecture (e.g. the classic by Tanenbaum) and read how computers really work. It's an entertaining read.

[andy_laycock]

"I recommend that you pick up a book on computer architecture (e.g. the classic by Tanenbaum) and read how computers really work. It's an entertaining read."

 

Did I say anything about computers? You're reading what you want into my comments.

[ilkka_nissila]

And what exactly is the difference between computers and digital electronics, from the point of view of analog modulation? Neither has any of it.

[ilkka_nissila]

(apart from analog-to-digital converters in computers)

 

Instead of taking on the excellent Tanenbaum book, if you are impatient you might just read

 

http://en.wikipedia.org/wiki/Digital_electronics

 

to see whether your views on digital vs. the brain are generally accepted. In fact virtually nothing mentioned on the page about digital electronics can be said about the brain!