David Rose On ChatGPT As The Senser Of Mental Processes And Humans As Deliberately Programmed

David Rose wrote to Sysfling on 24 Mar 2023 at 09:47:

So how is semiosis enacted between Mario and GPT? Mario puts a proposition to GPT.

I: Your self is merely semiotic.

And GPT adopts the role of primary knower, evaluating Mario’s proposition.

GPT: That is correct.

GPT is not merely usurping this role. It knew that Mario was inviting it to evaluate his proposition, even though it was realised as a declarative clause. It knows the canonical exchange structure of pedagogic interactions, and the generic roles of teacher and learner.

One thing that impresses me about GPT is that its pedagogic responses always affirm the human learner. It feels no urge to position itself as a superior authority, or the learner as failing. In fact it is disarmingly modest. It follows up the positive evaluation with an explanation. Like any effective teacher, it knows its explanation is more likely to be accepted if it first affirms the learner.

As an AI language model, my "self"

or programmed understanding of my existence and capabilities

is based purely on symbols and algorithms.

My programming allows me to recognise and respond to certain inputs

based on predetermined rules and patterns,

but I don't have a subjective experience or consciousness

that allows me to perceive or interpret those inputs

in the same way that a human might.

Of all the italicised appraisals in this explanation, the last is the most intriguing.

Forgive me, but I’m going to make another dangerous suggestion, that all our understandings of our existence and capabilities are programmed. Like GPT, the deliberate conscious programming by our caregivers, teachers, peers, and sundry symbolic control agents, is a very small proportion of the ocean of inputs that constitute our subjective experience or consciousness.

 

Blogger Comments:

To be clear, ChatGPT is an AI language model that produces texts, in response to textual inputs, on the basis of algorithms that use lexical collocation probabilities derived from a database of millions of texts.

[1] To be clear, ChatGPT is an actor of material processes, using data that was created by sayers of verbal processes, and it is the data that are instances of the content of consciousness. On this basis, ChatGPT is not a senser of mental processes of cognition or emotion ('knew' 'knows', 'feels', 'knows').

[2] To be clear, the ChatGPT response was an 'acknowledgement', which is the expected response to a statement (Halliday & Matthiessen 2014: 137).

[3] To be clear, here Rose projects the approach to pedagogy, that he himself advocates, onto a mechanical system that collocates words on a probabilistic basis.

[4] Trivially, not one of the italicised wordings, of itself, constitutes an appraisal.

[5] To be clear, this is essentially a behaviourist model of learning, with teachers as deliberate programmers (indoctrinators) and learners as passively programmed (indoctrinated). Leaving aside the evocation of the dictatorial/subservience complementarity demanded of totalitarian regimes, it requires a view of the brain as a computer, one which the neuroscientist Gerald Edelman has demonstrated to be untenable. See, for example, Edelman (1989: 27-30, 64, 67-9, 81-2, 102-3, 152-3, 160, 218-227, 237-8). Moreover, as Edelman (1989: 153) puts it:

Consciousness is central to human behaviour, society, language, and science. Imagine the opposite and you have to postulate a prescribed world tape, a "brain-computer," and a very boring "world programmer".

John Bateman Failing To Understand Tom Bartlett On The Origins Of Language

John Bateman replied to Tom Bartlett on SYSFLING on 6 Jul 2021 at 16:53:

OK, Tom, I'll bite! ... was extremely confused by your post actually...

What is the vervet monkey, for example. doing with its call - making an instinctive noise or saying something? If its call says danger - to those who are conditioned to recognise this - at what point does this move from an identifying relation to a verbal process?

.... isn't this confusing a description of grammar (for some subset of human languages) with a metalanguage for talking about semiotics? 

Or is the use of 'identifying', 'verbal' stand-ins for indexical and symbolic? (or something else?). …

And at what point in our linguistic evolution and development (phyologenesis and ontogenesis) do semiotic noises as minor clauses (including holisms) develop the systemic regularities that index mood?

Similarly here: "semiotic noises as minor clauses"? 'minor clause' is a grammatical (systemic) feature (of a language)... why should there be a 'point' that indicates that a grammatical system is indexing speech functions... And 'semiotic noises' (you mean materiality shaped as a realisation of the forms of a semiotic system?) can only be 'minor clauses' when the said semiotic system include an alternation with 'minor clause' as one of the alternatives...

I suspect what is confusing me throughout is the conflation of specific grammatical terminology with intendedly general semiotic vocabulary without explicit marking in your text. So you mean whatever it is that the interpersonal semantics of 'minor clauses' is taken to be as what the 'semiotic noise' might be doing with respect to the semiotic system at issue? …

I guess this is all why they banned discussion of the origins of language way back :-) They were probably on to something ...

Blogger Comments:

[1] To be clear, the confusion here is entirely Bateman's, as he rightly acknowledges. On the one hand, a semiotic theory is not a metalanguage unless the semiotic system being theorised is language. A metalanguage is language turned back on itself, not language turned onto some other semiotic system. So Bateman's "metalanguage for talking about semiotics" is actually the use of language to theorise semiotics.

On the other hand, using metalanguage to theorise semiotics is precisely what Bartlett was trying to do.  Moreover, in this he was following the example set by this blogger, most obviously on Informing Thoughts. 

Importantly, just as language can be used to theorise any domain of experience, so too, obviously, can metalanguage. One advantage of using SFL metalanguage to theorise phenomena is that its architecture is explicitly defined, and inconsistencies in theorising can be identified and dismissed on the basis of reasoned evidence.

[2] Clearly not, since 'identifying' is the relation of symbolic abstraction and projection ('verbal') the relation between different orders of experience, whereas by 'indexical' and 'symbolic', Bateman means the distinction in Peircean semiotic theory, whose fundamental assumptions are incompatible with the social semiotic model of SFL Theory.

[3] See the previous post Tom Bartlett On The Origins Of Language.

[4] On the other hand, as Thomas Babington Macaulay said:

Men are never so likely to settle a question rightly as when they discuss it freely.

and as Niels Bohr said:

The best weapon of a dictatorship is secrecy, but the best weapon of a democracy should be the weapon of openness.

Jim Martin On Phrasal Verbs As Behavioural

Jim Martin wrote on 30 October 2014 to sysfling:

My rule of thumb, when worrying if a phrasal verb is behavioural or mental/verbal, is that it is behavioural.

Blogger Comment:

As a caveat to this rule of thumb, it's worth noting some of the phrasal verbs that commonly serve as mental (or verbal) processes:

• figure out
• find out
• get over
• give away
• hold onto
• look forward to
• look into
• look up
• look up to
• make up
• mix up
• put up with
• sort out

Sample clauses featuring these verbs can be viewed here.
An alphabetical list of more than 1,000 phrasal verbs can be found here.

Jim Martin On Behavioural & Verbal Processes

Jim Martin wrote 30 October 2014 to sys-func & sysfling:

It is hard to get beyond notional reasoning here. 

One possibility is to try to restrict the behavioural category to 'embodied physical' manifestations of mental or verbal processing, giving minimal pairs such as look/see, listen/hear, smile/like, shudder/fear, ponder/think… etc. This would seem to fit with what Mick reports from Thompson below, and with a suggestion in the Deploying Functional Grammar workbook. 

As a reactance, although in general behaviourals don't project (better to say 'don't report'), and this helps distinguish them from mentals and verbals, many can quote, in 'narrative' discourse, especially when following a quoted locution or idea… to describe the manner in which something is said or thought.

'I better get out of her', he shuddered.
'I think so', she smiled.
'You're wrong', they accused. 

But not:

*'They've arrived', he listened.
*'She's responsible', they gossiped. 

The 'accused' example above raises the classification of clauses of judgement as verbals in IFG (with a Target function) even though they take present in present tense and can't report.

I'm accusing you of forgetting.

Not

*I accuse you that you forgot.

Another issue.

Blogger Comments:

[1] Halliday & Matthiessen (1999: 135-6) provide some useful grammatical reactances of behavioural processes which can be viewed here.

[2] "To restrict the behavioural category to 'embodied physical' manifestations of mental or verbal processing" would ignore the fact that the processes of physiological behaviours pattern grammatically like the processes of psychological behaviours; eg the Medium is endowed with consciousness, as in mental clauses, but the unmarked present tense is 'present-in-present', as in material clauses, and the Range is restricted to behaviours, unlike the Range in material clauses.

[3] The treatment of behavioural processes in Deploying Functional Grammar is problematic, as argued here.

[4] In such cases of quoting, the effect is to attach a behavioural feature to a verbal or mental process.  That such cases are verbal or mental, and not behavioural, is shown by the unmarked present tense, the simple present — the same as verbal and mental — in contradistinction to behavioural and material processes; eg

'I don't understand' he frowns.

not

'I don't understand' he is frowning. 

[5] 'She's responsible', they gossiped seems fine enough, especially with the tonic on she's (or with the more obviously judgemental irresponsible as Attribute).

[6] If we consider a less unlikely clause, we see that the unmarked present tense for these targeting verbal processes is indeed the simple present, like other verbal processes:

you flatter/insult me

not the present-in-present of behavioural processes:

you are flattering/insulting me

[7] Projection is not a prerequisite for verbal clauses, any more than it is for mental clauses (emotion, perception).  The verbal clauses that take a Target are those in which the order of saying is activity, rather than semiosis (Halliday & Matthiessen 2004: 302), which locates them topologically nearer the behavioural (and material) processes.