Mike, I suggest you go to *their* list and enlighten them on what they haven't thought through as well as you have... maybe they will show more interest than you get here.
Personally, although I agree that the authors you refer to (Lakoff, Johnson, etc) -- and many others directly in the field of Cognitive Linguistics as well as related fields -- are extremely useful for insights applicable to AGI research. Most of my own reading these days is in that area. Your own take on it seems bizarrely focused on one subpart of one sense modality, as if you skimmed a few articles and fixated on the word "image" which you could bend to your own preconceived ideas about the mind, which seem rather dubious. When Lakoff pats you on the head and refers you to source material, predictably you ignore the referenced work's substance and take the whole interaction as some sort of vindication.... Schemes (and image schemes) are much more abstract than you ever acknowledge... An IS for, say, COMPULSION, is not some kind of neural line drawing of arrows and balls... But AI has never really been about copying The brain (at any level of abstraction). It is astonishing and unbelievably cool that evolution, which produced a complex set of task-specific information processing mechanisms to solve survival problems, was then able to repurpose these to achieve (relatively) general purpose abstract cognition. But... To AGI folks, it would be equally astonishing if the result turned out to be the ONLY way intelligence can be built, or even the BEST way to build it using radically different technological substrates. It does seem to me that the resulting hodgepodge -- biological inspiration combined with intuitions about abstract reasoning -- is rather unconvincing without either firm theoretical foundations or clear incremental experimental progress... but that doesn't doom such efforts, it just makes the success of any particular effort rather unlikely. Your criticisms miss the mark here because you presume a requirement that AGI must operate the same way as your beliefs about how human minds operate, and your resulting anthropocentric view produces rather irrelevant comments about how AGI development MUST proceed. Evolution may have (roughly) started with generality and added sophistication and eventually "intelligence" to that, but it's not the only possibly way to proceed. It would be very interesting to see an AGI effort taking explicit direction from cognitive linguistics (especially fom a developmental perspective) and related areas of study focusing on concepts, schemas, and metaphorical mappings -- again especially from a perspective of "non-traditional" computational models... If I were to get anywhere in my own efforts, it would follow such a path. But it is far easier said than done. I don't know what that computational model should be (yet, haha), and neither as far as I know does anybody else including Lakoff and his colleagues. Their computational models of abstracted neural function are just as arbitrary and unconvincing as the mechanisms chosen by AGI researchers, and although they are better motivated for the purpose of a cognitive psychology research agenda, whether such parochial methods are best for AGI is much less clear, especially while the results are still so underwhelming. While I am rambling, I personally think that the vast pre-general-intelligence neural heritage provides a very rich many-dimensional sensorimotor environment on which abstract cognition rests. Thinking in terms, for example, of vision as a sort of movie camera is not the best way to think about it... Hardwired neural circuitry to separate form from movement, figure from ground, and objects from a spatial field (including slots for tracking those objects)... Spatial maps of various kinds; various spatial coordinate systems, etc, etc --- THOSE are what should properly be thought of as the outputs of sense modalities, and those are the things that schemas are built from, are the genesis of abstraction. These color our thinking and communication so much that from a purely practical standpoint I don't see how a "young" AGI could get much from reading human text, or communicate anything except the simplest technical material, without replicating a lot of that architectural detail. But that's just me... Derek Zahn From: [email protected] To: [email protected] Subject: Re: [agi] Image schemas control all forms of action [Lakoff replies] Date: Tue, 24 Jul 2012 10:10:22 +0100 The idea of mapping schemas onto each other is v. fundamental to the general image schema approach to cognition and action of cog. embodied sci. This is esp. evident in the whole idea of metaphors wh. is of extreme (and IMO somewhat excessive) importance to the field. Computers as we have much discussed aren’t yet capable of holistic mapping – though I think a way round can be found for robots. The other aspect of schemas that is vital is that they are fluid,loose outlines – and not just outlines of objects, but of actions and potential courses of actions – and therefore a fundamental contradiction and challenge to the idea of algorithmic, precisely first-to-last-step preplanned courses of action. *However* I doubt that anyone in the field has really thought this through – or they wouldn’t be so attracted to computational instantiations.. I welcome comments from all here. From: Jim Bromer Sent: Tuesday, July 24, 2012 2:43 AM To: AGI Subject: Re: [agi] Image schemas control all forms of action [Lakoff replies] Mike said: My impression is that these attempts are always misguided ... – for they do IMO “betray” or certainly distort the guiding image schema inspiration – and the idea of mapping schemas onto each other. (I’d like to discuss this with him/them – and may use your reply as an opportunity). Do you mean that the idea of mapping schema onto each other is a distortion of "the guiding image schema inspiration," or do you mean that the idea of mapping schema onto each other is a part of the inspiration? Because the "mapping" of schema onto each other is a mapping of a computational idea onto a hypothesis about the biological process of mind, whereas the guiding inspiration of using image is more of a process derived from physical neural biology. Jim Bromer On Mon, Jul 23, 2012 at 7:13 PM, Mike Tintner <[email protected]> wrote: Ben, Did you read it in the proper order, so to speak (hard to do from the layout)? i.e. starting with *my* post and his reply? I don’t think there’s any doubt that he is replying to, and confirming my position – wh. is a general point about how the brain works, and how image schemas inform and control many different kinds of action, incl. cognition and representation. It’s true that at almost every point, Lakoff and his many followers/colleagues seek to find computational instantiations of their ideas. My impression is that these attempts are always misguided – and invite the kind of response you have made, – for they do IMO “betray” or certainly distort the guiding image schema inspiration – and the idea of mapping schemas onto each other. (I’d like to discuss this with him/them – and may use your reply as an opportunity). But I don’t think there can be any doubt that Lakoff & co do see image schemas as central as I have outlined (and don’t see them as mathematical) – and that while they may seek to be computational, their primary loyalty is to the biological and science. From: Ben Goertzel Sent: Monday, July 23, 2012 11:54 PM To: AGI Cc: AGI Subject: Re: [agi] Image schemas control all forms of action [Lakoff replies] Mike, Lakoff's reply to you is not about "image schema" but rather about "process schema" , specifically naranyan's x-schema naranyan's x-schema are "a graph-based, token-passing formalism based on stochastic Petri nets" http://www1.icsi.berkeley.edu/~snarayan/CFN-NCPW04.pdf These x-schema are an abstract mathematical formalism, and not intrinsically "imagistic" Naranyan uses x-schema as a bridge btw language, action, perception and reasoning -- much as opencog uses its atomspace model in this role Ben G -- Ben Goertzel http://goertzel.org ### Sent from my mobile; plz forgive any typos or excessive concision ... On 24 Jul, 2012, at 5:17 AM, "Mike Tintner" <[email protected]> wrote: From: George Lakoff Sent: Monday, July 23, 2012 10:11 PM To: Mike Tintner Subject: Re: [Cogling-L] The scope of image schemas Narayanan's X-schemas (or process schemas) characterize all events and actions and actually control physical actions. So you're right about that. We are now working on entity schemas, but we're not there yet. George On Sun, Jul 22, 2012 at 11:34 AM, Mike Tintner <[email protected]> wrote: Lakoff:The idea behind image metaphors is simple. Images are structured by image schemas. A given image has multiple image schemas linked via neural binding to form a composite image schema ? or more than one. Metaphors map one image to another by mapping the source image schemas to the identical image schemas in the target George, Your exposition was v. useful. Can you/should you not extend the scope of image schemas? They structure presumably under *Images* : both *Verbal Images* & *Graphic/Photographic/Sensory Images*. and not just word images but : *Words/Language/Concepts" - period; *all words* are structured by image schemas, no? And from that one can one go on to argue - no? - that they structure *Moves/Movement* - period - that, for example, our reaching for a cup is structured by a schema. After all, language is used principally to structure actions: "Hand me that cup" - "Go to the other room". It makes sense that image schemas should structure not just verbally-mediated action, but all action, however mediated. The same mirror neurons that respond to (image-schema-structured) verbal accounts of action, also respond when just watching direct sensory images of agents executing those actions. Concepts/schemas arguably structure all the actions of living creatures. Comments? P.S. Personally, I think it's helpful to think of image schemas as "[loose] outlines" - esp. in connection with actions. Comments? AGI | Archives | Modify Your Subscription AGI | Archives | Modify Your Subscription AGI | Archives | Modify Your Subscription AGI | Archives | Modify Your Subscription AGI | Archives | Modify Your Subscription ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-c97d2393 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-2484a968 Powered by Listbox: http://www.listbox.com
