The hardware/software analogy of the BPS theory

[Snow Leopard]

Moderator note:
A series of posts has been split from the thread 'Protocol - Persistent symptoms Reduction Intervention - A system change and evaluation PRINCE, 2015 onwards, Chalder Moss-Morris et al'
******

And they’ve worded it in such a way so that people don’t look into it too much and worry. “Oh it’s ok I don’t actually have something wrong in “my actual physical body””.

The software/hardware analogy is indeed nonsense and it invokes the same mind-body dualism that they claim doesn't exist.

The rationale is explained here: "Explaining unexplained pain to fibromyalgia patients: finding a narrative that is acceptable to patients and provides a rationale for evidence based interventions"
https://journals.sagepub.com/doi/abs/10.1177/2049463716642601 (Hyland et al. 2016)

I don't think anyone who actually understands how biological systems or digital processing systems work would believe the analogy is reasonable. But I think it is more of a case of "whatever you say doctor", with patients humoring the therapist.

 

[rvallee]

I don't think anyone who actually understands how biological systems or digital processing systems work would believe the analogy is reasonable.

I'm a programmer and it's a dumb analogy that reflects ignorance of how computing works. It's not how any of this works.

The "software execution" in biological systems doesn't just flow in the hardware, it is hardware. It's chemical reactions, proteins folding, fibers stretching, gas exchange, electric potential and other very hardware stuff. The entire program is made of hardware and doesn't "run on hardware", it is hardware running itself.

Plus there is the fact that no matter how hard you try, you cannot change the function of or destroy the hardware using software alone (which is how they dismiss all the biological anomalies, as being influenced by the mind, somehow, because reasons). So it actually makes the opposite point and quite convincingly so. Technically you can execute a very demanding program (or just an endless loop, really) that sends the CPU into overdrive long enough to destroy some of its circuits (or probably the cooling fan before that) because of heat given enough time but that is something completely unique to computing.

 

[Kirsten]

Hi, I've had the computer analogy used on me a few times. At the time it just struck me as a dumb platitude. But I'd like to understand more about why you said it's not even a good analogy? Forgive me, my computing knowledge is minimal! If it's all "hardware running hardware" then what is the "software"? (The things I think I know: hardware = physical computer, software = programs run on it e.g .Microsoft word ?) If you have the time and energy could someone explain it simply to me, why it's a bad analogy, as I'd love to get my head around this. Thank you. K

 

[Barry]

The "software execution" in biological systems doesn't just flow in the hardware, it is hardware. It's chemical reactions, proteins folding, fibers stretching, gas exchange, electric potential and other very hardware stuff. The entire program is made of hardware and doesn't "run on hardware", it is hardware running itself.

I'm also a software developer, and I do find the waters a bit muddy here. Yes it is all hardware in the end, but begs the question: why call anything software? To me it is about abstraction. What we call software is an abstracted presentation of what would otherwise be phenomenally complex patterns of hardware binary switches turning on and off. We might argue that music is really patterns of physical pressure waves through air - which it ultimately is - but that would be ignoring how the abstraction we call music has its own presentation to us.

 

[Snowdrop]

Yes. I particularly did not understand this bit . . .
(Also, @rvallee the simpler the better)

Plus there is the fact that no matter how hard you try, you cannot change the function of or destroy the hardware using software alone (which is how they dismiss all the biological anomalies, as being influenced by the mind, somehow, because reasons).

 

[rvallee]

Hi, I've had the computer analogy used on me a few times. At the time it just struck me as a dumb platitude. But I'd like to understand more about why you said it's not even a good analogy? Forgive me, my computing knowledge is minimal! If it's all "hardware running hardware" then what is the "software"? (The things I think I know: hardware = physical computer, software = programs run on it e.g .Microsoft word ?) If you have the time and energy could someone explain it simply to me, why it's a bad analogy, as I'd love to get my head around this. Thank you. K

There simply isn't any comparison between the two. It's just a forced metaphor. Like comparing apples and heavy metal music. You could run the same computer program using an integrated circuit, levers and pulleys or water flowing through canals as long as you adhere to a standard protocol. The only difference will be the speed at which you get an output. It's all just a series of simple steps going extremely fast.

You can run a million programs a billion times each and the hardware will not change. It is immutable from the day it was etched. None of its resources are spent or modified. The process of running a program uses the hardware but leaves it entirely unchanged, cannot modify in any way that relates to the program itself, only physical stress or heat can alter it in any way (or just smashing it with a hammer, your choice).

Biology changes all the time. Cells die, make physical connections to one another, exchange proteins and synthesize them, move nutrients and various molecules that perform work. Our thoughts move through biological "circuitry" that changes all the time and uses internal resources, exchange them and morph themselves. The cells are attacked, some of their internal structures may go awry, DNA replication may fail. It can also function mostly normally but be missing critical stuff, like electrolytes or some vitamins. So many things that simply have no analogue with computing.

I'd definitely explain it better without that damn brain fog...

 

[rvallee]

Yes. I particularly did not understand this bit . . .
(Also, @rvallee the simpler the better)

Running low on brain fumes so pretty sure I'm not making much sense but that itself is a good example. Computers run or they don't. There isn't all that weird balance of thousands of different things being moved, spent or altered inside.

The brain itself and how it thinks is a function of how it is physically structured and altered. Autonomic processes cause all sorts of molecules, proteins and the like to be physically created, moved and spent. Wetware creates wetware, itself the product of how the wetware is organized in the first place and itself in its present configuration because of how DNA code, a physical arrangement itself, lead to more physical processes to synthesize protein, use molecules, spend chemical reactions, etc.

There simply is no equivalent in computer hardware. You can run any combination of software any number of times and the hardware will remain exactly as it was on the day it was produced. They are both truly and fundamentally separate, so much so that you can easily translate software to run on entirely different hardware rather easily.

Probably not making much sense. Anyway the analogy is just wrong and it's hard to explain how two things aren't alike when there are thousands of ways they are completely different.

 

[Snowdrop]

Thanks @rvallee appreciate your efforts. I think your further explanation clarifies things for me.

Til nexttime . . .

 

[Snow Leopard]

I'm also a software developer, and I do find the waters a bit muddy here. Yes it is all hardware in the end, but begs the question: why call anything software?

Because people sat down and wrote "software", which ultimately a list of binary logic instructions telling a digital computer what to do in a linear manner.

Formal languages (including programming languages) are very different from our thoughts, or biological signalling networks.

Life is a process, it is constantly maintaining/recreating itself. There is no start or end, unlike computer programs. There is also a lot of fuzziness or imprecision which doesn't really exist in computer software. Which is to say, software crashes, whereas life is surprisingly resilient.

There is also a scale aspect that doesn't translate either.
chemistry->biochemistry->cell chemistry->intracellular interactions->tissue structures->organs->organ systems->bodies->small scale social structures and body-environmental interactions->societal interactions and larger scale environmental interactions->...

None of this translates well to the "levels of abstraction" in computer science which is merely a unidirectional abstraction. In our computers, there is no backwards interaction between the larger scale structures and the smaller scale structures.

If you want to invoke the analogy of "software" have to ask yourself, where is the place for a linear list of formal instructions in such a system?

DNA is notable for only coding for a bunch of tools, it says nothing explicit about how those tools are to be used.

What we call software is an abstracted presentation of what would otherwise be phenomenally complex patterns of hardware binary switches turning on and off.

Compared to even the simplest biological systems (eg a bacteria), the complexity of our computing systems is trivial.

 

[Sean]

The software/hardware analogy is indeed nonsense and it invokes the same mind-body dualism that they claim doesn't exist.

Just another of the apparently endless stream of internal contradictions in the BPS model.

 

[Jonathan Edwards]

My understanding of this is that the hardware/software distinction can be helpful but it has nothing to do with things being physical or conceptual. Hardware is the arrangement of the computing system that determines its input-output behaviour and that is built in at the time of manufacture. Software is any modification to that arrangement thereafter that guides the input-output relation in a way suited to a particular task.

The distinction is purely a pragmatic one and there are lots of examples of arrangements that do not fit easily into one or other. A Wordstar chip might come pre-inserted into a BBC B computer or you might have to buy it and fit it later. It was considered software because it was dedicated to particular task.

Another generalisation about the difference between hardware and software is that hardware tends to define the number and arrangement of internal units that can be either one way or another (off or on states). Software in contrast tends to involve setting some of those units to specific patterns of off or on. Even if software comes as a 'hard chip' it very often has to be booted up into working memory as a pattern of off and on states that influence responses to inputs.

I think there is an analogy with brains here in that we have a certain number of Brin cells with connection options, where reinforced connections might be on states and inhibited connections off states. The shift from on to off is then a physiological option but certain patterns of changes in connections might be enough to make one seriously ill. Epilepsy might be an example.

 

[large donner]

Someone tell me that a software virus cant completely fuck a computer. Even if an engineer can fix the virus he has to know what he's doing not just stand next to the computer telling it not to worry and to think positively. Otherwise it would be fucked forever.

Surely if they want to make an analogy between hardware and software they should consider the enormous amount of evidence that environmental issues like bacteria, viruses, and other environmental issues etc can cause life long diseases.

Perhaps they think HIV is just software that's introduced after the bodies hardware is in place so therefore harmless.

 

[Kirsten]

Thank you. My brain is not taking this in but I've saved your answers for next time I have a bit of concentration. Really appreciate you taking time to explain it to me. K

 

[Snow Leopard]

I think there is an analogy with brains here in that we have a certain number of Brin cells with connection options, where reinforced connections might be on states and inhibited connections off states. The shift from on to off is then a physiological option but certain patterns of changes in connections might be enough to make one seriously ill. Epilepsy might be an example.

Brains are not arrays of digital logic gates with a fixed network configuration.

Nerves themselves can die or regenerate, it is not merely the 'switching' (nerve firing) that is important.

Epilepsy causes structural changes, so I'm not sure the analogy holds there. (note: a seizure is a symptom, not an illness)

 

[Jonathan Edwards]

Brains are not arrays of digital logic gates with a fixed network configuration.

Nerves themselves can die or regenerate, it is not merely the 'switching' (nerve firing) that is important.

I don't think one has to say they are for there to be a useful analogy. Brains are arrays of integrators that maybe instantiate logic rather more directly than digital gates. The configurations are pretty fixed. I doubt cell birth and death is very important, other than in terms of going senile. Moreover my phone keeps updating its programming.

 

[rvallee]

Thanks @rvallee appreciate your efforts. I think your further explanation clarifies things for me.

Til nexttime . . .

I'm not sure it does but I appreciate the thought

 

[Barry]

Because people sat down and wrote "software", which ultimately a list of binary logic instructions telling a digital computer what to do in a linear manner.

Formal languages (including programming languages) are very different from our thoughts, or biological signalling networks.

Life is a process, it is constantly maintaining/recreating itself. There is no start or end, unlike computer programs. There is also a lot of fuzziness or imprecision which doesn't really exist in computer software. Which is to say, software crashes, whereas life is surprisingly resilient.

There is also a scale aspect that doesn't translate either.
chemistry->biochemistry->cell chemistry->intracellular interactions->tissue structures->organs->organ systems->bodies->small scale social structures and body-environmental interactions->societal interactions and larger scale environmental interactions->...

None of this translates well to the "levels of abstraction" in computer science which is merely a unidirectional abstraction. In our computers, there is no backwards interaction between the larger scale structures and the smaller scale structures.

If you want to invoke the analogy of "software" have to ask yourself, where is the place for a linear list of formal instructions in such a system?

DNA is notable for only coding for a bunch of tools, it says nothing explicit about how those tools are to be used.



Compared to even the simplest biological systems (eg a bacteria), the complexity of our computing systems is trivial.

If I say to someone who has just exchanged contracts for their house, "you've burnt your boats now", I think we all know I'm not being literal. An analogy is about illustrating a similarity between things that are otherwise likely extremely different. I most certainly don't think our brains and their processing actually operate the same as computers and software - obviously they are hugely different.

If you want to invoke the analogy of "software" have to ask yourself, where is the place for a linear list of formal instructions in such a system?

Depends on the analogy being drawn. Never crossed my mind our brains might process everything in linear fashion, quite the opposite I've always assumed. It's just an analogy - the hierarchical layers - that has grown with me over my 30+ years writing embedded firmware as well as software, long before ever hearing of PR, S4ME or even ME.

... However ... as I write this I think I can see where you and others are coming from, and can agree with it. The mechanism that underpins information processing in our brains has evolved, and as such it is only the physical elements that can have done so - there is no higher entity 'writing' a more abstracted form of it; it is what it is.

But I do think analogies should be considered within their intended context, and not automatically criticised for not being literal enough.

 

[Barry]

There simply isn't any comparison between the two. It's just a forced metaphor. Like comparing apples and heavy metal music. You could run the same computer program using an integrated circuit, levers and pulleys or water flowing through canals as long as you adhere to a standard protocol. The only difference will be the speed at which you get an output. It's all just a series of simple steps going extremely fast.

You can run a million programs a billion times each and the hardware will not change. It is immutable from the day it was etched. None of its resources are spent or modified. The process of running a program uses the hardware but leaves it entirely unchanged, cannot modify in any way that relates to the program itself, only physical stress or heat can alter it in any way (or just smashing it with a hammer, your choice).

Biology changes all the time. Cells die, make physical connections to one another, exchange proteins and synthesize them, move nutrients and various molecules that perform work. Our thoughts move through biological "circuitry" that changes all the time and uses internal resources, exchange them and morph themselves. The cells are attacked, some of their internal structures may go awry, DNA replication may fail. It can also function mostly normally but be missing critical stuff, like electrolytes or some vitamins. So many things that simply have no analogue with computing.

I'd definitely explain it better without that damn brain fog...

I assumed the analogy was about hardware-control software, rather than about software somehow magically modifying the hardware itself. I've written stacks of firmware and software that controls hardware and monitors what it is doing; indeed the hardware drivers installed on our PCs are doing exactly that as we use them. If software could not control hardware then it would mostly be for nothing. Maybe when one person is talking about software controlling hardware, another person is interpreting that as software changing the hardware. Controlling hardware is about influencing its behaviour, and software can only do that if the hardware has been designed to that end. Such hardware will not be changed by the software, but its behaviour can be.

 

[Adrian]

My understanding of this is that the hardware/software distinction can be helpful but it has nothing to do with things being physical or conceptual. Hardware is the arrangement of the computing system that determines its input-output behaviour and that is built in at the time of manufacture. Software is any modification to that arrangement thereafter that guides the input-output relation in a way suited to a particular task.

I don't think that the analogy is a good one (and not good for computers either).

At the hardware layer what we really have is a series of transistors (and other electrical components). A chip is made up of transistors which are arranged in a particular way to compute a given function (or algorithm) then there are layers above this that compute more complex algorithms. So there may be an arrangement of gates or transistors which form an arithmetic pipeline which computes basic numeric computations other gates would control what they do and thus we can build up more complex algorithms. Some will be computed due to the arrangement of the gates but others will be done as 'software' which is basically changeable rather than fixed. But which is which is just an implementation decision around the complexity and flexibility of the compute surface. Something like an FPGA allows hardware gates to be programmed and reprogrammed with a given algorithm and that can be quicker than a general purpose chip but it does demonstrate some general concepts. I think what is important is the concepts of the gates, the algorithms and some notion of programability ( in terms of the ability to change algorithms) - I think memory is also a concept which should come in here.

I think a further concept is important which is that of a distributed system. When most talk about distributed systems they refer to things running on multiple computers but we can consider a PE with say a web cam and a disk as a distributed system. The disk has hardware and software and does the job of writing blocks to a platter (or a flash memory). The Webcam has hardware and software which manages how the images are taken and filtered. Both will be connected via a bus (PCI or USB) to the main processor. In fact even the main processor can be thought of as a distributed system in that it has multiple interconnected cores each running different code along with other components such as graphics functions and network functions. The basic concept of importance it that the processing (both gates and algorithms can be spread around and the results connected).

I think there is an analogy with brains here in that we have a certain number of Brin cells with connection options, where reinforced connections might be on
states and inhibited connections off states. The shift from on to off is then a physiological option but certain patterns of changes in connections might be enough to make one seriously ill. Epilepsy might be an example.

We could argue that the analogy is between gates with simple functions and brain cells but I suspect they are more complex that that and compute some more complex function which is also dependent on the chemicals around in the brain (gates in a chip work differently if subjected to the wrong voltage or a fault attack). Then I suspect there are layers of algorithms and memories distributed over a the neurons as a compute surface - this would fit with neural network research as well as brain imaging suggesting local activation being associated with different concepts. But I also suspect that the layers of algorithms will build up complexity so one may spot movements on a retinal image and another may convert that into a predictor/pray signal where as another algorithm in a different area of the brain would reason about actions etc.

 

[Jonathan Edwards]

I don't disagree with any of that @Adrian but I don't see in what way it is at odds with what I said.

Neurons certainly have much more complex ways of computing outputs. That is what I work on when I am not thinking about ME. They can probably evaluate complex truths as opposed to computing that 1=1. They can indicate near matching and compete to indicate the best matching, and so on.