Oregan-2<= /TITLE> <META content=3D"text/html; charset=3Dwindows-1252" = http-equiv=3DContent-Type> <META name=3DProgId content=3DWord.Document> <META name=3DGENERATOR content=3D"MSHTML 8.00.7600.16490"> <META name=3DOriginator content=3D"Microsoft Word 11"><LINK = rel=3DFile-List=20 href=3D"Rogers-McClelland_files/filelist.xml"><LINK rel=3DEdit-Time-Data = href=3D"Rogers-McClelland_files/editdata.mso"><o:SmartTagType name=3D"PostalCode"=20 namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"></o:SmartTagT= ype><o:SmartTagType=20 name=3D"State"=20 namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"></o:SmartTagT= ype><o:SmartTagType=20 name=3D"City"=20 namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"></o:SmartTagT= ype><o:SmartTagType=20 name=3D"place"=20 namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"></o:SmartTagT= ype><o:SmartTagType=20 name=3D"PlaceType"=20 namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"></o:SmartTagT= ype><o:SmartTagType=20 name=3D"PlaceName"=20 namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"></o:SmartTagT= ype> <STYLE>@font-face { font-family: Tahoma; } @font-face { font-family: Cambria; } @page Section1 {size: 8.5in 11.0in; margin: 1.0in 1.25in 1.0in 1.25in; = mso-header-margin: .5in; mso-footer-margin: .5in; mso-paper-source: 0; } P.MSONORMAL { mso-style-unhide: no; mso-style-qformat: yes } LI.MSONORMAL { mso-style-unhide: no; mso-style-qformat: yes } DIV.MSONORMAL { mso-style-unhide: no; mso-style-qformat: yes } H1 { mso-style-unhide: no; mso-style-qformat: yes; mso-fareast-theme-font: = minor-fareast } H2 { mso-style-unhide: no; mso-style-qformat: yes; mso-fareast-theme-font: = minor-fareast } P.MSOBODYTEXT { mso-style-unhide: no } LI.MSOBODYTEXT { mso-style-unhide: no } DIV.MSOBODYTEXT { mso-style-unhide: no } A:link { mso-style-unhide: no } SPAN.MSOHYPERLINK { mso-style-unhide: no } A:visited { mso-style-unhide: no } SPAN.MSOHYPERLINKFOLLOWED { mso-style-unhide: no } P { mso-style-unhide: no } ADDRESS { mso-style-unhide: no; mso-fareast-theme-font: minor-fareast } P.MSOACETATE { mso-style-unhide: no } LI.MSOACETATE { mso-style-unhide: no } DIV.MSOACETATE { mso-style-unhide: no } SPAN.HTMLADDRESSCHAR { mso-style-unhide: no } SPAN.HEADING1CHAR { mso-style-unhide: no; mso-fareast-theme-font: major-fareast; = mso-ascii-theme-font: major-latin; mso-hansi-theme-font: major-latin; = mso-bidi-theme-font: major-bidi; mso-themecolor: accent1; = mso-themeshade: 191 } SPAN.HEADING2CHAR { mso-style-unhide: no; mso-fareast-theme-font: major-fareast; = mso-ascii-theme-font: major-latin; mso-hansi-theme-font: major-latin; = mso-bidi-theme-font: major-bidi; mso-themecolor: accent1 } SPAN.BODYTEXTCHAR { mso-style-unhide: no } SPAN.BALLOONTEXTCHAR { mso-style-unhide: no } .MSOCHPDEFAULT { mso-default-props: yes } TABLE.MSONORMALTABLE { mso-style-qformat: yes; mso-style-priority: 99 } P.MSONORMAL { MARGIN: 0in 0in 0pt; FONT-FAMILY: "Times New Roman"; COLOR: black; = FONT-SIZE: 12pt; mso-style-parent: ""; mso-pagination: widow-orphan; = mso-fareast-font-family: "Times New Roman" } LI.MSONORMAL { MARGIN: 0in 0in 0pt; FONT-FAMILY: "Times New Roman"; COLOR: black; = FONT-SIZE: 12pt; mso-style-parent: ""; mso-pagination: widow-orphan; = mso-fareast-font-family: "Times New Roman" } DIV.MSONORMAL { MARGIN: 0in 0in 0pt; FONT-FAMILY: "Times New Roman"; COLOR: black; = FONT-SIZE: 12pt; mso-style-parent: ""; mso-pagination: widow-orphan; = mso-fareast-font-family: "Times New Roman" } H1 { FONT-FAMILY: "Times New Roman"; COLOR: black; MARGIN-LEFT: 0in; = FONT-SIZE: 24pt; FONT-WEIGHT: bold; MARGIN-RIGHT: 0in; mso-pagination: = widow-orphan; mso-style-link: "Heading 1 Char"; mso-margin-top-alt: = auto; mso-margin-bottom-alt: auto; mso-outline-level: 1 } H2 { FONT-FAMILY: "Times New Roman"; COLOR: black; MARGIN-LEFT: 0in; = FONT-SIZE: 18pt; FONT-WEIGHT: bold; MARGIN-RIGHT: 0in; mso-pagination: = widow-orphan; mso-style-link: "Heading 2 Char"; mso-margin-top-alt: = auto; mso-margin-bottom-alt: auto; mso-outline-level: 2 } P.MSOBODYTEXT { MARGIN: 0in 0in 0pt; FONT-FAMILY: Arial; COLOR: windowtext; FONT-SIZE: = 11pt; mso-pagination: widow-orphan; mso-fareast-font-family: "Times New = Roman"; mso-style-link: "Body Text Char"; mso-style-noshow: yes; = mso-bidi-font-size: 10.0pt; mso-bidi-font-family: "Times New Roman" } LI.MSOBODYTEXT { MARGIN: 0in 0in 0pt; FONT-FAMILY: Arial; COLOR: windowtext; FONT-SIZE: = 11pt; mso-pagination: widow-orphan; mso-fareast-font-family: "Times New = Roman"; mso-style-link: "Body Text Char"; mso-style-noshow: yes; = mso-bidi-font-size: 10.0pt; mso-bidi-font-family: "Times New Roman" } DIV.MSOBODYTEXT { MARGIN: 0in 0in 0pt; FONT-FAMILY: Arial; COLOR: windowtext; FONT-SIZE: = 11pt; mso-pagination: widow-orphan; mso-fareast-font-family: "Times New = Roman"; mso-style-link: "Body Text Char"; mso-style-noshow: yes; = mso-bidi-font-size: 10.0pt; mso-bidi-font-family: "Times New Roman" } A:link { COLOR: blue; TEXT-DECORATION: underline; text-underline: single } SPAN.MSOHYPERLINK { COLOR: blue; TEXT-DECORATION: underline; text-underline: single } A:visited { COLOR: blue; TEXT-DECORATION: underline; text-underline: single } SPAN.MSOHYPERLINKFOLLOWED { COLOR: blue; TEXT-DECORATION: underline; text-underline: single } P { FONT-FAMILY: "Times New Roman"; COLOR: black; MARGIN-LEFT: 0in; = FONT-SIZE: 12pt; MARGIN-RIGHT: 0in; mso-pagination: widow-orphan; = mso-fareast-font-family: "Times New Roman"; mso-margin-top-alt: auto; = mso-margin-bottom-alt: auto } ADDRESS { FONT-STYLE: italic; MARGIN: 0in 0in 0pt; FONT-FAMILY: "Times New = Roman"; COLOR: black; FONT-SIZE: 12pt; mso-pagination: widow-orphan; = mso-style-link: "HTML Address Char" } P.MSOACETATE { MARGIN: 0in 0in 0pt; FONT-FAMILY: Tahoma; COLOR: black; FONT-SIZE: 8pt; = mso-pagination: widow-orphan; mso-fareast-font-family: "Times New = Roman"; mso-style-link: "Balloon Text Char" } LI.MSOACETATE { MARGIN: 0in 0in 0pt; FONT-FAMILY: Tahoma; COLOR: black; FONT-SIZE: 8pt; = mso-pagination: widow-orphan; mso-fareast-font-family: "Times New = Roman"; mso-style-link: "Balloon Text Char" } DIV.MSOACETATE { MARGIN: 0in 0in 0pt; FONT-FAMILY: Tahoma; COLOR: black; FONT-SIZE: 8pt; = mso-pagination: widow-orphan; mso-fareast-font-family: "Times New = Roman"; mso-style-link: "Balloon Text Char" } SPAN.HTMLADDRESSCHAR { FONT-STYLE: italic; COLOR: black; mso-style-link: "HTML Address"; = mso-bidi-font-size: 12.0pt; mso-style-name: "HTML Address Char"; = mso-style-locked: yes; mso-ansi-font-size: 12.0pt } SPAN.HEADING1CHAR { FONT-FAMILY: Cambria; COLOR: #365f91; FONT-WEIGHT: bold; = mso-fareast-font-family: "Times New Roman"; mso-style-link: "Heading 1"; = mso-bidi-font-size: 14.0pt; mso-bidi-font-family: "Times New Roman"; = mso-style-name: "Heading 1 Char"; mso-style-locked: yes; = mso-ansi-font-size: 14.0pt; mso-ascii-font-family: Cambria; = mso-hansi-font-family: Cambria } SPAN.HEADING2CHAR { FONT-FAMILY: Cambria; COLOR: #4f81bd; FONT-WEIGHT: bold; = mso-fareast-font-family: "Times New Roman"; mso-style-link: "Heading 2"; = mso-bidi-font-size: 13.0pt; mso-bidi-font-family: "Times New Roman"; = mso-style-name: "Heading 2 Char"; mso-style-locked: yes; = mso-ansi-font-size: 13.0pt; mso-ascii-font-family: Cambria; = mso-hansi-font-family: Cambria } SPAN.BODYTEXTCHAR { COLOR: black; mso-style-link: "Body Text"; mso-bidi-font-size: 12.0pt; = mso-style-name: "Body Text Char"; mso-style-locked: yes; = mso-ansi-font-size: 12.0pt } SPAN.BALLOONTEXTCHAR { FONT-FAMILY: Tahoma; COLOR: black; mso-style-link: "Balloon Text"; = mso-bidi-font-size: 8.0pt; mso-bidi-font-family: Tahoma; mso-style-name: = "Balloon Text Char"; mso-style-locked: yes; mso-ansi-font-size: 8.0pt; = mso-ascii-font-family: Tahoma; mso-hansi-font-family: Tahoma } SPAN.SpellE { mso-style-name: ""; mso-spl-e: yes } SPAN.GramE { mso-style-name: ""; mso-gram-e: yes } DIV.Section1 { page: Section1 } </STYLE> <LINK rel=3DthemeData=20 href=3D"Rogers-McClelland_files/themedata.thmx"><LINK = rel=3DcolorSchemeMapping=20 href=3D"Rogers-McClelland_files/colorschememapping.xml"> <META name=3DTemplate content=3D"C:\PROGRAM FILES\MICROSOFT = OFFICE\OFFICE\html.dot"></HEAD> <BODY style=3D"tab-interval: .5in" lang=3DEN-US aLink=3D#0000ff = link=3Dblue=20 bgColor=3Dwhite vLink=3Dblue> <DIV class=3DSection1> <A=20 href=3D"http://www.bbsonline.org/"><SPAN=20 style=3D"FONT-SIZE: 36pt">BBS<SPAN=20 style=3D"FONT-SIZE: 36pt; TEXT-DECORATION: none; text-underline: none; = mso-no-proof: yes"><IMG=20 id=3D_x0000_i1025 border=3D0 src=3D"" width=3D49 = height=3D49><SPAN=20 style=3D"FONT-SIZE: 36pt">nline</A> <SPAN=20 style=3D"FONT-SIZE: 14pt">Commentary on <I=20 style=3D"mso-bidi-font-style: normal">Semantic Cognition, by Rogers = and=20 McClelland <o:p></o:p> Abstract: <SPAN=20 style=3D"mso-tab-count: 1"> = 65=20 words Main Text: = 1355=20 words References: =20 385 words Total Text: =20 1805 words <H1>Semantic Redintegration: Ecological=20 Invariance</H1> <ADDRESS><st1:PlaceName w:st=3D"on">Stephen</st1:PlaceName> = <st1:PlaceName=20 w:st=3D"on">E.</st1:PlaceName> <st1:PlaceName=20 w:st=3D"on">Robbins</st1:PlaceName> <st1:PlaceType=20 w:st=3D"on">Center</st1:PlaceType> for Advanced Product Engineering, Metavante Corporation, <st1:place = w:st=3D"on"><st1:City=20 w:st=3D"on">Milwaukee</st1:City>, <st1:State w:st=3D"on">WI</st1:State>, = <st1:PostalCode w:st=3D"on">53224</st1:PostalCode></st1:place></ADDRESS> <ADDRESS><o:p> </o:p></ADDRESS> <ADDRESS>Home: 262-628-3502</ADDRESS> <ADDRESS>Homepage: =20 stephenerobbins.com</ADDRESS> <ADDRESS> <A=20 href=3D"mailto:Stephen.Robbins@metavante.com">mailto:Stephen.Robbins@meta= vante.com</A></ADDRESS> <H2>Abstract</H2> <SPAN=20 style=3D"FONT-FAMILY: 'Times New Roman'; FONT-SIZE: 12pt; = mso-bidi-font-size: 10.0pt">In=20 proposing that their model can operate in the concrete, perceptual = world, the=20 authors have not done justice to the complexities of the ecological = sphere and=20 its invariance laws. The=20 structure of concrete events forces a different framework, both for = retrieval of=20 events and concepts defined across events, than that upon which the = proposed=20 model, rooted in essence in the verbal learning tradition, implicitly=20 rests. =20 <o:p></o:p> ---------------------------- There is no cognition without memory, that is, without = the <SPAN=20 class=3DSpellE>redintegration of events.<SPAN=20 style=3D"mso-spacerun: yes"> The depth and breadth of = thought by=20 Rogers and McClelland on semantics is impressive, but I fear the claim = that=20 their model dwells in the perceptual/ecological sphere ignores the = realities of=20 this sphere, in the end resulting in an inadequate theory of <SPAN=20 class=3DSpellE>redintegration. =20 <o:p> </o:p> Consider an event: stirring coffee in a cup, using a = spoon.<SPAN=20 style=3D"mso-spacerun: yes"> The event has a time-extended = invariance=20 structure, here defined as the transformations and invariants specifying = an=20 event and rendering it a virtual action. =20 The swirling coffee surface is a radial flow field.<SPAN=20 style=3D"mso-spacerun: yes"> The constant size of the cup, = should it=20 move forward or backward, is specified, over time, by a constant ratio = of height=20 to the occluded texture units of the table surface gradient.<SPAN=20 style=3D"mso-spacerun: yes"> The tau ratio=20 defined over this flow field supports modulating the hand for grasping = the cup=20 (Savelsburgh et. al, 1991).<SPAN=20 style=3D"mso-spacerun: yes"> Were the cup cubical, its = edges and=20 vertices are sharp discontinuities in the velocity flows of its sides as = the=20 eyes saccade, where these flows specify the form of the cup (cf. = Robbins, 2004,=20 2007). The periodic = motion of the=20 spoon is a haptic flow field that carries an = adiabatic=20 invariance =96 a constant ratio of energy of oscillation to frequency of = oscillation (Kugler and <SPAN=20 class=3DSpellE>Turvey, 1987). =20 The action of wielding the spoon is defined by an inertial tensor = describing the moments of force (Turvey and = <SPAN=20 class=3DSpellE>Carello, 1995). =20 This entire informational structure and far more must be = supported,=20 globally, over time, by the resonant feedback among visual, motor, = auditory,=20 even prefrontal areas. To store such a dynamic, ever changing, global = pattern=20 entirely in the hippocampus for future feeding to the cortex, as = suggested=20 (McClelland et al., 1995), is an impossibility.<SPAN=20 style=3D"mso-spacerun: yes"> <o:p> </o:p> In the authors=92 formulation, I present an = occurrence, SPOON, in=20 the context of CAN, and the network must be trained, via weight = adjustments, to=20 respond with STIR. Or equivalently, in the causal version, the network = is=20 trained to predict the sequelae, e.g., the = =93circular=20 motion of the coffee liquid,=94 or the clinking sound, etc. What sense = does this=20 make? This is a remnant = of the=20 supremely non-ecological verbal learning tradition, its roots in the=20 semantics-eradication program of Ebbinghaus, = which=20 ultimately bifurcated events arbitrarily into components, e.g., SPOON = and=20 COFFEE, then asked: how do we learn these components as a = paired-associate=20 pair? This is the = PA-learning=20 framework of the older cousin-model (McClelland et al., 1995).<SPAN=20 style=3D"mso-spacerun: yes"> Paivio=92s=20 (1971) introduction of imagery, and later the elaboration techniques, = where=20 SPOON and COFFEE are imagined in a dynamic event, were the first = somewhat=20 ecological cracks in this brute force, syntactic learning framework. <o:p> </o:p> In reality, we are perceiving the spoon as an integral = part of a=20 stirring event, with all the event=92s ongoing invariance = structure.<SPAN=20 style=3D"mso-spacerun: yes"> Where is the =93error?=94 We need no weight adjustments = to =93link=94=20 the event =93components.=94 =20 This is precisely why the mechanism of back propagation or its = equivalent=20 will never be found. A = spoon=20 scooping and lifting oatmeal is yet another event with an integral and = complex=20 set of forces and auditory/visual patterning.<SPAN=20 style=3D"mso-spacerun: yes"> A spoon stirring pancake = batter is=20 yet a different complex invariance structure.<SPAN=20 style=3D"mso-spacerun: yes"> A spoon digging into and = cutting and=20 lifting grapefruit is yet another. =20 A spoon balancing on the edge of the coffee cup another.<SPAN=20 style=3D"mso-spacerun: yes"> Now we have the = set:<SPAN=20 style=3D"mso-spacerun: yes"> SPOON CAN: [stir, balance, = scoop, cut].=20 If I re-present SPOON, which event will it <SPAN=20 class=3DSpellE>redintegrate? =20 Presenting SPOON is roughly equivalent to a static event, a = resting=20 spoon. There is little = structure to=20 this event as a cue; it is underspecified, and yet is common to = all.<SPAN=20 style=3D"mso-spacerun: yes"> It is like sending an = imprecise=20 reconstructive wave with little coherence through a hologram =96 we = reconstruct a=20 composite image of multiple recorded wave fronts, or in this case, = stirring=20 events. It is the classic = interference of McGeoch (1942).<SPAN=20 style=3D"mso-spacerun: yes"> A network does not have to be=20 error-trained to specify this set. <o:p> </o:p> But equally, we can retrieve a specific event.<SPAN=20 style=3D"mso-spacerun: yes"> The cue must bear the same = invariance=20 structure or a sufficient subset. =20 For the coffee stirring, we re-present an abstract rendering of = the=20 coffee=92s radial flow field, or simulate the inertial tensor of the=20 wielding. We are = creating, globally=20 throughout the brain, a more constrained, more coherent = =93reconstructive=20 wave.=94 To reconstruct = the=20 batter-stirring event as opposed to the coffee-stirring, we must = constrain our=20 wielding cue differently to capture the larger amplitude of the stirring = motion=20 and/or the greater resistance of the medium.<SPAN=20 style=3D"mso-spacerun: yes"> In essence, we are discussing = a =93paired=20 associate=94 paradigm, call it A-Bi, that is impossible in = the verbal=20 learning realm, for every cue is in effect SPOON, yet with the = appropriate=20 ecological, dynamic constraints on the cue event involving a spoon, we = can=20 reconstruct each separate event (Robbins, 2006a, 2006b).<SPAN=20 style=3D"mso-spacerun: yes"> But<SPAN=20 style=3D"mso-spacerun: yes"> we also understand, = =93A SPOON can=20 CATAPULT (a pea),=94 for the spoon, we know, can be inserted in, and = supports the=20 forces/invariance structure of, catapulting.<SPAN=20 style=3D"mso-spacerun: yes"> This is precisely the = realm of=20 French=92s (1990, 1999; cf. Robbins, 2002) devastating critique re the = Turing=20 Test, and the network proposed is no better equipped for these emergent=20 analogies. <o:p> </o:p> What now is an end element of the =93hierarchy,=94 = e.g., the concept=20 STIR? For = coffee-stirring, it is a=20 higher order invariance existing or taken across many instances of the = dynamic=20 invariance structure described above. =20 Similar to Goldinger=92s (1998) = =93exemplar=94=20 suggestion, it would be the firing of the family of all =93traces=94 or=20 pattern/paths ever supporting a coffee-stirring event.<SPAN=20 style=3D"mso-spacerun: yes"> The result is a context free=20 COFFEE-STIRRING, while STIRRING in general is a higher order invariance = over a=20 yet larger number of event patterns, to include stirring cake batters, = soups,=20 coffee and cement. = This=20 is the meaning of the abstract, general, context-free STIRRING, but this = meaning=20 is intrinsically based in a system that can equally reconstruct the = specific,=20 e.g., the precise event of the cake-batter stirring versus the coffee = stirring,=20 via cueing by any appropriately constrained, sufficient subset of the = invariance=20 structure of these events. = <o:p> </o:p> Such a view questions the book=92s model of memory at = its=20 base. For relating = arbitrary,=20 unrelated events, we perhaps need a syntactic association mechanism of = some=20 form. For ecological = events with=20 their invariance structures, the principle of retrieval, for both = individual=20 events and abstract classes defined across events, is quite = different.<SPAN=20 style=3D"mso-spacerun: yes"> <o:p> </o:p> Balancing a scale is an event. To derive the full = semantics of=20 the principle of torque in and across such events, there is again the = perception=20 of invariance over transformation. =20 We specify the successive balancing transformations re the two = sides: 2=20 lbs placed at 10 inches (from center) or 2 lbs x 10 inches =3D 5 lbs x 4 = inches; 3=20 lbs x 6.66 inches =3D 5 lbs x 4 inches; 4 lbs x 5 inches =3D 5 lbs x 4 = inches; 6 lbs=20 x 3.33 inches =3D 5 lbs x 4 inches. =20 Over this transformation set, the invariance law emerges.<SPAN=20 style=3D"mso-spacerun: yes"> This invariance is the = semantics of=20 torque. Quinlan et al. = (2007) have=20 shown that in balance scale simulations, connectionist networks never = learn the=20 principle of torque, nor do many internal representations of approximate = rules=20 correspond to the human. = There is a=20 significance to this, deeply related to the subject =96 semantic = cognition =96 and=20 it is that connectionist models are insensitive to the transformations = and=20 invariants defining events. = It=20 would be an easy exercise to show that this will be true for nearly all = <SPAN=20 class=3DSpellE>Piagetian tasks. =20 What else is the tunnel-bead experiment (Piaget, 1946), which = demands the=20 child predict the oscillation of order in three colored beads moved into = a=20 cylinder/tunnel, given any n semi-rotations of the tunnel? <o:p> </o:p> Discovering invariance laws, it has been heavily = argued (<SPAN=20 class=3DSpellE>Kugler and Turvey, = 1987; <SPAN=20 class=3DSpellE>Wigner, 1970; Woit, = 2006;=20 Woodward, 2000; 2001), is = scientific=20 explanation. This = subsumes the=20 =93causal=94 explanations of theory-theory. =20 In this, science models itself after the brain in perception, and = in this=20 sense the authors, in holding, contra theory-theory, that there is no = special=20 mechanism, are perceptively right. =20 But coherent covariation is a low = order,=20 syntactic form of invariance. The neural dynamics required for the=20 transformations and invariance noted above, I believe, is beyond the = coherent=20 covariation detection of the model; the = lesson of=20 invariance in the ecological sphere has yet to be engaged by either = side.<SPAN=20 style=3D"mso-spacerun: yes"> = <B=20 style=3D"mso-bidi-font-weight: normal"><o:p></o:p> <H2>References</H2> <P=20 style=3D"TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; mso-hyphenate: none; = tab-stops: -.5in"=20 class=3DMsoNormal><SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">French, R. M. = (1990).=20 Sub-cognition and the limits of the Turing Test. <I=20 style=3D"mso-bidi-font-style: normal">Mind, <I=20 style=3D"mso-bidi-font-style: normal">99, = 53-65.<o:p></o:p> <P=20 style=3D"TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; mso-hyphenate: none; = tab-stops: -.5in"=20 class=3DMsoNormal><SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">French, R. M. = (1999). When=20 coffee cups are like old elephants, or why representation modules = don=92t make=20 sense. In <I=20 style=3D"mso-bidi-font-style: normal">Understanding Representation in = the=20 Cognitive Sciences, eds., A. Riegler, M. = <SPAN=20 class=3DSpellE>Peshl, & A. von Stein. = Plenum.<o:p></o:p> <P=20 style=3D"TEXT-ALIGN: justify; TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; = mso-hyphenate: none; tab-stops: -.5in"=20 class=3DMsoNormal><SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: = 12.0pt">Goldinger<SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">, S. (1998). = Echoes of=20 echoes? An episodic = theory of=20 lexical access. Psychological = Review,=20 105, = 251-279.<o:p></o:p> <SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: = 12.0pt">Kugler<SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">, P., & <SPAN=20 class=3DSpellE>Turvey, M. (1987). Information, Natural Law, = and the=20 Self-assembly of Rhythmic Movement. =20 Erlbaum.<o:p></o:p> <SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">McClelland, J. L., = <SPAN=20 class=3DSpellE>McNaughton, B. L., & O'Reilly, R. C. (1995). = Why there=20 are complementary learning systems in the hippocampus and <SPAN=20 class=3DSpellE>neocortex: = Insights from the successes and failures of connectionist models = of=20 learning and memory. <I=20 style=3D"mso-bidi-font-style: normal">Psychological Review, <I=20 style=3D"mso-bidi-font-style: normal">103, = 419-457.<o:p></o:p> <SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: = 12.0pt">McGeoch<SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">, J. A.<SPAN=20 style=3D"mso-spacerun: yes"> (1942). <I=20 style=3D"mso-bidi-font-style: normal">The Psychology of Human = Learning.<SPAN=20 style=3D"mso-spacerun: yes"> Longmans, = Greene.<o:p></o:p> <SPAN=20 style=3D"FONT-SIZE: 11pt">Paivio,=20 A. (1971).<SPAN=20 style=3D"mso-spacerun: yes"> <I=20 style=3D"mso-bidi-font-style: normal">Imagery and Verbal Processes. = Holt,=20 Rinehart, and Winston.<o:p></o:p> <P=20 style=3D"TEXT-ALIGN: justify; TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; = mso-hyphenate: none; tab-stops: -.5in"=20 class=3DMsoNormal>Piaget, J. (1946). <I=20 style=3D"mso-bidi-font-style: normal">The Child's Conception of Movement = and=20 Speed. <SPAN=20 style=3D"FONT-SIZE: 11pt; mso-ansi-language: DE"=20 lang=3DDE>Ballentine.<o:p></o:p> <P=20 style=3D"TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; mso-hyphenate: none; = tab-stops: -.5in"=20 class=3DMsoNormal>Quinlan, P., van der=20 Maas, H., Jansen, B., Booij, O., = Rendell, M. (2007). =20 Re-thinking stages of cognitive development: An appraisal = of=20 connectionist models of the balance scale task.<SPAN=20 style=3D"mso-spacerun: yes"> <I=20 style=3D"mso-bidi-font-style: normal">Cognition, <I=20 style=3D"mso-bidi-font-style: normal">103, 413-459. <SPAN=20 style=3D"FONT-SIZE: 11pt">Robbins, S. E. (2002). Semantics, experience = and=20 time. <I=20 style=3D"mso-bidi-font-style: normal">Cognitive Systems Research, 3, 301-335. = <o:p></o:p> <SPAN=20 style=3D"FONT-SIZE: 11pt">Robbins, S. E. (2004). On time, memory and = dynamic=20 form. <I=20 style=3D"mso-bidi-font-style: normal">Consciousness and Cognition, = <I=20 style=3D"mso-bidi-font-style: normal">13, 762-788.<SPAN=20 style=3D"mso-spacerun: yes"> <o:p></o:p> <SPAN=20 style=3D"FONT-SIZE: 11pt">Robbins, S. E. (2006a). On the possibility of = direct=20 memory. In New = Developments in=20 Consciousness Research, <SPAN=20 style=3D"mso-bidi-font-style: italic">ed.,V. W. = <SPAN=20 class=3DSpellE>Fallio. Nova Science, 1-64. <o:p></o:p> Robbins, S. E. = (2006b). <SPAN=20 class=3DSpellE>Bergson and the holographic theory of mind.<SPAN=20 style=3D"mso-spacerun: yes"> <I=20 style=3D"mso-bidi-font-style: normal">Phenomenology and the Cognitive=20 Sciences, 5,=20 365-394.<o:p></o:p> Robbins, S. E. (2007). = Time, form=20 and the limits of qualia. <I=20 style=3D"mso-bidi-font-style: normal">Journal of Mind and Behavior, = <I=20 style=3D"mso-bidi-font-style: normal">28, = 19-43.<o:p></o:p> <SPAN=20 class=3DSpellE><SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: = 12.0pt">Savelsbergh<SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">, G. J. P., = Whiting, H.T.,=20 & Bootsma, R. J. (1991).<SPAN=20 style=3D"mso-spacerun: yes"> Grasping <SPAN=20 class=3DSpellE>tau. = <I=20 style=3D"mso-bidi-font-style: normal">Journal of Experimental = Psychology: Human=20 Perception and Performance, 17,=20 315-322.<o:p></o:p> <P=20 style=3D"TEXT-ALIGN: justify; TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; = mso-hyphenate: none; tab-stops: -.5in"=20 class=3DMsoNormal><SPAN=20 style=3D"LETTER-SPACING: -0.15pt; FONT-SIZE: = 11pt">Turvey<SPAN=20 style=3D"LETTER-SPACING: -0.15pt; FONT-SIZE: 11pt">, M., & <SPAN=20 class=3DSpellE>Carello, C. (1995). Dynamic touch.<SPAN=20 style=3D"mso-spacerun: yes"> In <I=20 style=3D"mso-bidi-font-style: normal">Perception of Space and = Motion, eds., W.=20 Epstein & S. Rogers. = Academic=20 Press. = <o:p></o:p> <SPAN=20 style=3D"FONT-FAMILY: 'Times New Roman'">Wigner<SPAN=20 style=3D"FONT-FAMILY: 'Times New Roman'">, E. P. (1970).<SPAN=20 style=3D"mso-spacerun: yes"> Symmetries and = Reflections.<SPAN=20 style=3D"mso-spacerun: yes"> MIT = Press.<o:p></o:p> <SPAN=20 style=3D"FONT-FAMILY: 'Times New Roman'">Woit<SPAN=20 style=3D"FONT-FAMILY: 'Times New Roman'">, P. (2006).<SPAN=20 style=3D"mso-spacerun: yes"> <I=20 style=3D"mso-bidi-font-style: normal">Not Even Wrong: The Failure of = String Theory=20 and the Search for Unity in Physical Law.<SPAN=20 style=3D"mso-spacerun: yes"> Basic = Books.<o:p></o:p> <P=20 style=3D"TEXT-ALIGN: justify; TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; = mso-hyphenate: none; tab-stops: -.5in"=20 class=3DMsoNormal><SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">Woodward, J. = (2000).=20 Explanation and invariance in the special sciences.<SPAN=20 style=3D"mso-spacerun: yes"> British Journal for the = Philosophy of=20 Science, 51, = 197-214.<SPAN=20 style=3D"mso-spacerun: yes"> <o:p></o:p> <P=20 style=3D"TEXT-ALIGN: justify; TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; = mso-hyphenate: none; tab-stops: -.5in"=20 class=3DMsoNormal><SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">Woodward, J. = (2001). Law and=20 explanation in biology: Invariance is the kind of stability that = matters.<SPAN=20 style=3D"mso-spacerun: yes"> Philosophy of Science, = <I=20 style=3D"mso-bidi-font-style: normal">68, = 1-20.<o:p></o:p> . </DIV></BODY></HTML>

From: "Saved by Windows Internet Explorer 8" Subject: Oregan-2 Date: Fri, 26 Mar 2010 17:50:20 -0500 MIME-Version: 1.0 Content-Type: text/html; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable Content-Location: http://www.stephenerobbins.com/Articles/Robbins-Rogers-McClelland.html X-MimeOLE: Produced By Microsoft MimeOLE V6.1.7600.16385 Oregan-2<= /TITLE> <META content=3D"text/html; charset=3Dwindows-1252" = http-equiv=3DContent-Type> <META name=3DProgId content=3DWord.Document> <META name=3DGENERATOR content=3D"MSHTML 8.00.7600.16490"> <META name=3DOriginator content=3D"Microsoft Word 11"><LINK = rel=3DFile-List=20 href=3D"Rogers-McClelland_files/filelist.xml"><LINK rel=3DEdit-Time-Data = href=3D"Rogers-McClelland_files/editdata.mso"><o:SmartTagType name=3D"PostalCode"=20 namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"></o:SmartTagT= ype><o:SmartTagType=20 name=3D"State"=20 namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"></o:SmartTagT= ype><o:SmartTagType=20 name=3D"City"=20 namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"></o:SmartTagT= ype><o:SmartTagType=20 name=3D"place"=20 namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"></o:SmartTagT= ype><o:SmartTagType=20 name=3D"PlaceType"=20 namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"></o:SmartTagT= ype><o:SmartTagType=20 name=3D"PlaceName"=20 namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"></o:SmartTagT= ype> <STYLE>@font-face { font-family: Tahoma; } @font-face { font-family: Cambria; } @page Section1 {size: 8.5in 11.0in; margin: 1.0in 1.25in 1.0in 1.25in; = mso-header-margin: .5in; mso-footer-margin: .5in; mso-paper-source: 0; } P.MSONORMAL { mso-style-unhide: no; mso-style-qformat: yes } LI.MSONORMAL { mso-style-unhide: no; mso-style-qformat: yes } DIV.MSONORMAL { mso-style-unhide: no; mso-style-qformat: yes } H1 { mso-style-unhide: no; mso-style-qformat: yes; mso-fareast-theme-font: = minor-fareast } H2 { mso-style-unhide: no; mso-style-qformat: yes; mso-fareast-theme-font: = minor-fareast } P.MSOBODYTEXT { mso-style-unhide: no } LI.MSOBODYTEXT { mso-style-unhide: no } DIV.MSOBODYTEXT { mso-style-unhide: no } A:link { mso-style-unhide: no } SPAN.MSOHYPERLINK { mso-style-unhide: no } A:visited { mso-style-unhide: no } SPAN.MSOHYPERLINKFOLLOWED { mso-style-unhide: no } P { mso-style-unhide: no } ADDRESS { mso-style-unhide: no; mso-fareast-theme-font: minor-fareast } P.MSOACETATE { mso-style-unhide: no } LI.MSOACETATE { mso-style-unhide: no } DIV.MSOACETATE { mso-style-unhide: no } SPAN.HTMLADDRESSCHAR { mso-style-unhide: no } SPAN.HEADING1CHAR { mso-style-unhide: no; mso-fareast-theme-font: major-fareast; = mso-ascii-theme-font: major-latin; mso-hansi-theme-font: major-latin; = mso-bidi-theme-font: major-bidi; mso-themecolor: accent1; = mso-themeshade: 191 } SPAN.HEADING2CHAR { mso-style-unhide: no; mso-fareast-theme-font: major-fareast; = mso-ascii-theme-font: major-latin; mso-hansi-theme-font: major-latin; = mso-bidi-theme-font: major-bidi; mso-themecolor: accent1 } SPAN.BODYTEXTCHAR { mso-style-unhide: no } SPAN.BALLOONTEXTCHAR { mso-style-unhide: no } .MSOCHPDEFAULT { mso-default-props: yes } TABLE.MSONORMALTABLE { mso-style-qformat: yes; mso-style-priority: 99 } P.MSONORMAL { MARGIN: 0in 0in 0pt; FONT-FAMILY: "Times New Roman"; COLOR: black; = FONT-SIZE: 12pt; mso-style-parent: ""; mso-pagination: widow-orphan; = mso-fareast-font-family: "Times New Roman" } LI.MSONORMAL { MARGIN: 0in 0in 0pt; FONT-FAMILY: "Times New Roman"; COLOR: black; = FONT-SIZE: 12pt; mso-style-parent: ""; mso-pagination: widow-orphan; = mso-fareast-font-family: "Times New Roman" } DIV.MSONORMAL { MARGIN: 0in 0in 0pt; FONT-FAMILY: "Times New Roman"; COLOR: black; = FONT-SIZE: 12pt; mso-style-parent: ""; mso-pagination: widow-orphan; = mso-fareast-font-family: "Times New Roman" } H1 { FONT-FAMILY: "Times New Roman"; COLOR: black; MARGIN-LEFT: 0in; = FONT-SIZE: 24pt; FONT-WEIGHT: bold; MARGIN-RIGHT: 0in; mso-pagination: = widow-orphan; mso-style-link: "Heading 1 Char"; mso-margin-top-alt: = auto; mso-margin-bottom-alt: auto; mso-outline-level: 1 } H2 { FONT-FAMILY: "Times New Roman"; COLOR: black; MARGIN-LEFT: 0in; = FONT-SIZE: 18pt; FONT-WEIGHT: bold; MARGIN-RIGHT: 0in; mso-pagination: = widow-orphan; mso-style-link: "Heading 2 Char"; mso-margin-top-alt: = auto; mso-margin-bottom-alt: auto; mso-outline-level: 2 } P.MSOBODYTEXT { MARGIN: 0in 0in 0pt; FONT-FAMILY: Arial; COLOR: windowtext; FONT-SIZE: = 11pt; mso-pagination: widow-orphan; mso-fareast-font-family: "Times New = Roman"; mso-style-link: "Body Text Char"; mso-style-noshow: yes; = mso-bidi-font-size: 10.0pt; mso-bidi-font-family: "Times New Roman" } LI.MSOBODYTEXT { MARGIN: 0in 0in 0pt; FONT-FAMILY: Arial; COLOR: windowtext; FONT-SIZE: = 11pt; mso-pagination: widow-orphan; mso-fareast-font-family: "Times New = Roman"; mso-style-link: "Body Text Char"; mso-style-noshow: yes; = mso-bidi-font-size: 10.0pt; mso-bidi-font-family: "Times New Roman" } DIV.MSOBODYTEXT { MARGIN: 0in 0in 0pt; FONT-FAMILY: Arial; COLOR: windowtext; FONT-SIZE: = 11pt; mso-pagination: widow-orphan; mso-fareast-font-family: "Times New = Roman"; mso-style-link: "Body Text Char"; mso-style-noshow: yes; = mso-bidi-font-size: 10.0pt; mso-bidi-font-family: "Times New Roman" } A:link { COLOR: blue; TEXT-DECORATION: underline; text-underline: single } SPAN.MSOHYPERLINK { COLOR: blue; TEXT-DECORATION: underline; text-underline: single } A:visited { COLOR: blue; TEXT-DECORATION: underline; text-underline: single } SPAN.MSOHYPERLINKFOLLOWED { COLOR: blue; TEXT-DECORATION: underline; text-underline: single } P { FONT-FAMILY: "Times New Roman"; COLOR: black; MARGIN-LEFT: 0in; = FONT-SIZE: 12pt; MARGIN-RIGHT: 0in; mso-pagination: widow-orphan; = mso-fareast-font-family: "Times New Roman"; mso-margin-top-alt: auto; = mso-margin-bottom-alt: auto } ADDRESS { FONT-STYLE: italic; MARGIN: 0in 0in 0pt; FONT-FAMILY: "Times New = Roman"; COLOR: black; FONT-SIZE: 12pt; mso-pagination: widow-orphan; = mso-style-link: "HTML Address Char" } P.MSOACETATE { MARGIN: 0in 0in 0pt; FONT-FAMILY: Tahoma; COLOR: black; FONT-SIZE: 8pt; = mso-pagination: widow-orphan; mso-fareast-font-family: "Times New = Roman"; mso-style-link: "Balloon Text Char" } LI.MSOACETATE { MARGIN: 0in 0in 0pt; FONT-FAMILY: Tahoma; COLOR: black; FONT-SIZE: 8pt; = mso-pagination: widow-orphan; mso-fareast-font-family: "Times New = Roman"; mso-style-link: "Balloon Text Char" } DIV.MSOACETATE { MARGIN: 0in 0in 0pt; FONT-FAMILY: Tahoma; COLOR: black; FONT-SIZE: 8pt; = mso-pagination: widow-orphan; mso-fareast-font-family: "Times New = Roman"; mso-style-link: "Balloon Text Char" } SPAN.HTMLADDRESSCHAR { FONT-STYLE: italic; COLOR: black; mso-style-link: "HTML Address"; = mso-bidi-font-size: 12.0pt; mso-style-name: "HTML Address Char"; = mso-style-locked: yes; mso-ansi-font-size: 12.0pt } SPAN.HEADING1CHAR { FONT-FAMILY: Cambria; COLOR: #365f91; FONT-WEIGHT: bold; = mso-fareast-font-family: "Times New Roman"; mso-style-link: "Heading 1"; = mso-bidi-font-size: 14.0pt; mso-bidi-font-family: "Times New Roman"; = mso-style-name: "Heading 1 Char"; mso-style-locked: yes; = mso-ansi-font-size: 14.0pt; mso-ascii-font-family: Cambria; = mso-hansi-font-family: Cambria } SPAN.HEADING2CHAR { FONT-FAMILY: Cambria; COLOR: #4f81bd; FONT-WEIGHT: bold; = mso-fareast-font-family: "Times New Roman"; mso-style-link: "Heading 2"; = mso-bidi-font-size: 13.0pt; mso-bidi-font-family: "Times New Roman"; = mso-style-name: "Heading 2 Char"; mso-style-locked: yes; = mso-ansi-font-size: 13.0pt; mso-ascii-font-family: Cambria; = mso-hansi-font-family: Cambria } SPAN.BODYTEXTCHAR { COLOR: black; mso-style-link: "Body Text"; mso-bidi-font-size: 12.0pt; = mso-style-name: "Body Text Char"; mso-style-locked: yes; = mso-ansi-font-size: 12.0pt } SPAN.BALLOONTEXTCHAR { FONT-FAMILY: Tahoma; COLOR: black; mso-style-link: "Balloon Text"; = mso-bidi-font-size: 8.0pt; mso-bidi-font-family: Tahoma; mso-style-name: = "Balloon Text Char"; mso-style-locked: yes; mso-ansi-font-size: 8.0pt; = mso-ascii-font-family: Tahoma; mso-hansi-font-family: Tahoma } SPAN.SpellE { mso-style-name: ""; mso-spl-e: yes } SPAN.GramE { mso-style-name: ""; mso-gram-e: yes } DIV.Section1 { page: Section1 } </STYLE> <LINK rel=3DthemeData=20 href=3D"Rogers-McClelland_files/themedata.thmx"><LINK = rel=3DcolorSchemeMapping=20 href=3D"Rogers-McClelland_files/colorschememapping.xml"> <META name=3DTemplate content=3D"C:\PROGRAM FILES\MICROSOFT = OFFICE\OFFICE\html.dot"></HEAD> <BODY style=3D"tab-interval: .5in" lang=3DEN-US aLink=3D#0000ff = link=3Dblue=20 bgColor=3Dwhite vLink=3Dblue> <DIV class=3DSection1> <A=20 href=3D"http://www.bbsonline.org/"><SPAN=20 style=3D"FONT-SIZE: 36pt">BBS<SPAN=20 style=3D"FONT-SIZE: 36pt; TEXT-DECORATION: none; text-underline: none; = mso-no-proof: yes"><IMG=20 id=3D_x0000_i1025 border=3D0 src=3D"" width=3D49 = height=3D49><SPAN=20 style=3D"FONT-SIZE: 36pt">nline</A> <SPAN=20 style=3D"FONT-SIZE: 14pt">Commentary on <I=20 style=3D"mso-bidi-font-style: normal">Semantic Cognition, by Rogers = and=20 McClelland <o:p></o:p> Abstract: <SPAN=20 style=3D"mso-tab-count: 1"> = 65=20 words Main Text: = 1355=20 words References: =20 385 words Total Text: =20 1805 words <H1>Semantic Redintegration: Ecological=20 Invariance</H1> <ADDRESS><st1:PlaceName w:st=3D"on">Stephen</st1:PlaceName> = <st1:PlaceName=20 w:st=3D"on">E.</st1:PlaceName> <st1:PlaceName=20 w:st=3D"on">Robbins</st1:PlaceName> <st1:PlaceType=20 w:st=3D"on">Center</st1:PlaceType> for Advanced Product Engineering, Metavante Corporation, <st1:place = w:st=3D"on"><st1:City=20 w:st=3D"on">Milwaukee</st1:City>, <st1:State w:st=3D"on">WI</st1:State>, = <st1:PostalCode w:st=3D"on">53224</st1:PostalCode></st1:place></ADDRESS> <ADDRESS><o:p> </o:p></ADDRESS> <ADDRESS>Home: 262-628-3502</ADDRESS> <ADDRESS>Homepage: =20 stephenerobbins.com</ADDRESS> <ADDRESS> <A=20 href=3D"mailto:Stephen.Robbins@metavante.com">mailto:Stephen.Robbins@meta= vante.com</A></ADDRESS> <H2>Abstract</H2> <SPAN=20 style=3D"FONT-FAMILY: 'Times New Roman'; FONT-SIZE: 12pt; = mso-bidi-font-size: 10.0pt">In=20 proposing that their model can operate in the concrete, perceptual = world, the=20 authors have not done justice to the complexities of the ecological = sphere and=20 its invariance laws. The=20 structure of concrete events forces a different framework, both for = retrieval of=20 events and concepts defined across events, than that upon which the = proposed=20 model, rooted in essence in the verbal learning tradition, implicitly=20 rests. =20 <o:p></o:p> ---------------------------- There is no cognition without memory, that is, without = the <SPAN=20 class=3DSpellE>redintegration of events.<SPAN=20 style=3D"mso-spacerun: yes"> The depth and breadth of = thought by=20 Rogers and McClelland on semantics is impressive, but I fear the claim = that=20 their model dwells in the perceptual/ecological sphere ignores the = realities of=20 this sphere, in the end resulting in an inadequate theory of <SPAN=20 class=3DSpellE>redintegration. =20 <o:p> </o:p> Consider an event: stirring coffee in a cup, using a = spoon.<SPAN=20 style=3D"mso-spacerun: yes"> The event has a time-extended = invariance=20 structure, here defined as the transformations and invariants specifying = an=20 event and rendering it a virtual action. =20 The swirling coffee surface is a radial flow field.<SPAN=20 style=3D"mso-spacerun: yes"> The constant size of the cup, = should it=20 move forward or backward, is specified, over time, by a constant ratio = of height=20 to the occluded texture units of the table surface gradient.<SPAN=20 style=3D"mso-spacerun: yes"> The tau ratio=20 defined over this flow field supports modulating the hand for grasping = the cup=20 (Savelsburgh et. al, 1991).<SPAN=20 style=3D"mso-spacerun: yes"> Were the cup cubical, its = edges and=20 vertices are sharp discontinuities in the velocity flows of its sides as = the=20 eyes saccade, where these flows specify the form of the cup (cf. = Robbins, 2004,=20 2007). The periodic = motion of the=20 spoon is a haptic flow field that carries an = adiabatic=20 invariance =96 a constant ratio of energy of oscillation to frequency of = oscillation (Kugler and <SPAN=20 class=3DSpellE>Turvey, 1987). =20 The action of wielding the spoon is defined by an inertial tensor = describing the moments of force (Turvey and = <SPAN=20 class=3DSpellE>Carello, 1995). =20 This entire informational structure and far more must be = supported,=20 globally, over time, by the resonant feedback among visual, motor, = auditory,=20 even prefrontal areas. To store such a dynamic, ever changing, global = pattern=20 entirely in the hippocampus for future feeding to the cortex, as = suggested=20 (McClelland et al., 1995), is an impossibility.<SPAN=20 style=3D"mso-spacerun: yes"> <o:p> </o:p> In the authors=92 formulation, I present an = occurrence, SPOON, in=20 the context of CAN, and the network must be trained, via weight = adjustments, to=20 respond with STIR. Or equivalently, in the causal version, the network = is=20 trained to predict the sequelae, e.g., the = =93circular=20 motion of the coffee liquid,=94 or the clinking sound, etc. What sense = does this=20 make? This is a remnant = of the=20 supremely non-ecological verbal learning tradition, its roots in the=20 semantics-eradication program of Ebbinghaus, = which=20 ultimately bifurcated events arbitrarily into components, e.g., SPOON = and=20 COFFEE, then asked: how do we learn these components as a = paired-associate=20 pair? This is the = PA-learning=20 framework of the older cousin-model (McClelland et al., 1995).<SPAN=20 style=3D"mso-spacerun: yes"> Paivio=92s=20 (1971) introduction of imagery, and later the elaboration techniques, = where=20 SPOON and COFFEE are imagined in a dynamic event, were the first = somewhat=20 ecological cracks in this brute force, syntactic learning framework. <o:p> </o:p> In reality, we are perceiving the spoon as an integral = part of a=20 stirring event, with all the event=92s ongoing invariance = structure.<SPAN=20 style=3D"mso-spacerun: yes"> Where is the =93error?=94 We need no weight adjustments = to =93link=94=20 the event =93components.=94 =20 This is precisely why the mechanism of back propagation or its = equivalent=20 will never be found. A = spoon=20 scooping and lifting oatmeal is yet another event with an integral and = complex=20 set of forces and auditory/visual patterning.<SPAN=20 style=3D"mso-spacerun: yes"> A spoon stirring pancake = batter is=20 yet a different complex invariance structure.<SPAN=20 style=3D"mso-spacerun: yes"> A spoon digging into and = cutting and=20 lifting grapefruit is yet another. =20 A spoon balancing on the edge of the coffee cup another.<SPAN=20 style=3D"mso-spacerun: yes"> Now we have the = set:<SPAN=20 style=3D"mso-spacerun: yes"> SPOON CAN: [stir, balance, = scoop, cut].=20 If I re-present SPOON, which event will it <SPAN=20 class=3DSpellE>redintegrate? =20 Presenting SPOON is roughly equivalent to a static event, a = resting=20 spoon. There is little = structure to=20 this event as a cue; it is underspecified, and yet is common to = all.<SPAN=20 style=3D"mso-spacerun: yes"> It is like sending an = imprecise=20 reconstructive wave with little coherence through a hologram =96 we = reconstruct a=20 composite image of multiple recorded wave fronts, or in this case, = stirring=20 events. It is the classic = interference of McGeoch (1942).<SPAN=20 style=3D"mso-spacerun: yes"> A network does not have to be=20 error-trained to specify this set. <o:p> </o:p> But equally, we can retrieve a specific event.<SPAN=20 style=3D"mso-spacerun: yes"> The cue must bear the same = invariance=20 structure or a sufficient subset. =20 For the coffee stirring, we re-present an abstract rendering of = the=20 coffee=92s radial flow field, or simulate the inertial tensor of the=20 wielding. We are = creating, globally=20 throughout the brain, a more constrained, more coherent = =93reconstructive=20 wave.=94 To reconstruct = the=20 batter-stirring event as opposed to the coffee-stirring, we must = constrain our=20 wielding cue differently to capture the larger amplitude of the stirring = motion=20 and/or the greater resistance of the medium.<SPAN=20 style=3D"mso-spacerun: yes"> In essence, we are discussing = a =93paired=20 associate=94 paradigm, call it A-Bi, that is impossible in = the verbal=20 learning realm, for every cue is in effect SPOON, yet with the = appropriate=20 ecological, dynamic constraints on the cue event involving a spoon, we = can=20 reconstruct each separate event (Robbins, 2006a, 2006b).<SPAN=20 style=3D"mso-spacerun: yes"> But<SPAN=20 style=3D"mso-spacerun: yes"> we also understand, = =93A SPOON can=20 CATAPULT (a pea),=94 for the spoon, we know, can be inserted in, and = supports the=20 forces/invariance structure of, catapulting.<SPAN=20 style=3D"mso-spacerun: yes"> This is precisely the = realm of=20 French=92s (1990, 1999; cf. Robbins, 2002) devastating critique re the = Turing=20 Test, and the network proposed is no better equipped for these emergent=20 analogies. <o:p> </o:p> What now is an end element of the =93hierarchy,=94 = e.g., the concept=20 STIR? For = coffee-stirring, it is a=20 higher order invariance existing or taken across many instances of the = dynamic=20 invariance structure described above. =20 Similar to Goldinger=92s (1998) = =93exemplar=94=20 suggestion, it would be the firing of the family of all =93traces=94 or=20 pattern/paths ever supporting a coffee-stirring event.<SPAN=20 style=3D"mso-spacerun: yes"> The result is a context free=20 COFFEE-STIRRING, while STIRRING in general is a higher order invariance = over a=20 yet larger number of event patterns, to include stirring cake batters, = soups,=20 coffee and cement. = This=20 is the meaning of the abstract, general, context-free STIRRING, but this = meaning=20 is intrinsically based in a system that can equally reconstruct the = specific,=20 e.g., the precise event of the cake-batter stirring versus the coffee = stirring,=20 via cueing by any appropriately constrained, sufficient subset of the = invariance=20 structure of these events. = <o:p> </o:p> Such a view questions the book=92s model of memory at = its=20 base. For relating = arbitrary,=20 unrelated events, we perhaps need a syntactic association mechanism of = some=20 form. For ecological = events with=20 their invariance structures, the principle of retrieval, for both = individual=20 events and abstract classes defined across events, is quite = different.<SPAN=20 style=3D"mso-spacerun: yes"> <o:p> </o:p> Balancing a scale is an event. To derive the full = semantics of=20 the principle of torque in and across such events, there is again the = perception=20 of invariance over transformation. =20 We specify the successive balancing transformations re the two = sides: 2=20 lbs placed at 10 inches (from center) or 2 lbs x 10 inches =3D 5 lbs x 4 = inches; 3=20 lbs x 6.66 inches =3D 5 lbs x 4 inches; 4 lbs x 5 inches =3D 5 lbs x 4 = inches; 6 lbs=20 x 3.33 inches =3D 5 lbs x 4 inches. =20 Over this transformation set, the invariance law emerges.<SPAN=20 style=3D"mso-spacerun: yes"> This invariance is the = semantics of=20 torque. Quinlan et al. = (2007) have=20 shown that in balance scale simulations, connectionist networks never = learn the=20 principle of torque, nor do many internal representations of approximate = rules=20 correspond to the human. = There is a=20 significance to this, deeply related to the subject =96 semantic = cognition =96 and=20 it is that connectionist models are insensitive to the transformations = and=20 invariants defining events. = It=20 would be an easy exercise to show that this will be true for nearly all = <SPAN=20 class=3DSpellE>Piagetian tasks. =20 What else is the tunnel-bead experiment (Piaget, 1946), which = demands the=20 child predict the oscillation of order in three colored beads moved into = a=20 cylinder/tunnel, given any n semi-rotations of the tunnel? <o:p> </o:p> Discovering invariance laws, it has been heavily = argued (<SPAN=20 class=3DSpellE>Kugler and Turvey, = 1987; <SPAN=20 class=3DSpellE>Wigner, 1970; Woit, = 2006;=20 Woodward, 2000; 2001), is = scientific=20 explanation. This = subsumes the=20 =93causal=94 explanations of theory-theory. =20 In this, science models itself after the brain in perception, and = in this=20 sense the authors, in holding, contra theory-theory, that there is no = special=20 mechanism, are perceptively right. =20 But coherent covariation is a low = order,=20 syntactic form of invariance. The neural dynamics required for the=20 transformations and invariance noted above, I believe, is beyond the = coherent=20 covariation detection of the model; the = lesson of=20 invariance in the ecological sphere has yet to be engaged by either = side.<SPAN=20 style=3D"mso-spacerun: yes"> = <B=20 style=3D"mso-bidi-font-weight: normal"><o:p></o:p> <H2>References</H2> <P=20 style=3D"TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; mso-hyphenate: none; = tab-stops: -.5in"=20 class=3DMsoNormal><SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">French, R. M. = (1990).=20 Sub-cognition and the limits of the Turing Test. <I=20 style=3D"mso-bidi-font-style: normal">Mind, <I=20 style=3D"mso-bidi-font-style: normal">99, = 53-65.<o:p></o:p> <P=20 style=3D"TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; mso-hyphenate: none; = tab-stops: -.5in"=20 class=3DMsoNormal><SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">French, R. M. = (1999). When=20 coffee cups are like old elephants, or why representation modules = don=92t make=20 sense. In <I=20 style=3D"mso-bidi-font-style: normal">Understanding Representation in = the=20 Cognitive Sciences, eds., A. Riegler, M. = <SPAN=20 class=3DSpellE>Peshl, & A. von Stein. = Plenum.<o:p></o:p> <P=20 style=3D"TEXT-ALIGN: justify; TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; = mso-hyphenate: none; tab-stops: -.5in"=20 class=3DMsoNormal><SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: = 12.0pt">Goldinger<SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">, S. (1998). = Echoes of=20 echoes? An episodic = theory of=20 lexical access. Psychological = Review,=20 105, = 251-279.<o:p></o:p> <SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: = 12.0pt">Kugler<SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">, P., & <SPAN=20 class=3DSpellE>Turvey, M. (1987). Information, Natural Law, = and the=20 Self-assembly of Rhythmic Movement. =20 Erlbaum.<o:p></o:p> <SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">McClelland, J. L., = <SPAN=20 class=3DSpellE>McNaughton, B. L., & O'Reilly, R. C. (1995). = Why there=20 are complementary learning systems in the hippocampus and <SPAN=20 class=3DSpellE>neocortex: = Insights from the successes and failures of connectionist models = of=20 learning and memory. <I=20 style=3D"mso-bidi-font-style: normal">Psychological Review, <I=20 style=3D"mso-bidi-font-style: normal">103, = 419-457.<o:p></o:p> <SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: = 12.0pt">McGeoch<SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">, J. A.<SPAN=20 style=3D"mso-spacerun: yes"> (1942). <I=20 style=3D"mso-bidi-font-style: normal">The Psychology of Human = Learning.<SPAN=20 style=3D"mso-spacerun: yes"> Longmans, = Greene.<o:p></o:p> <SPAN=20 style=3D"FONT-SIZE: 11pt">Paivio,=20 A. (1971).<SPAN=20 style=3D"mso-spacerun: yes"> <I=20 style=3D"mso-bidi-font-style: normal">Imagery and Verbal Processes. = Holt,=20 Rinehart, and Winston.<o:p></o:p> <P=20 style=3D"TEXT-ALIGN: justify; TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; = mso-hyphenate: none; tab-stops: -.5in"=20 class=3DMsoNormal>Piaget, J. (1946). <I=20 style=3D"mso-bidi-font-style: normal">The Child's Conception of Movement = and=20 Speed. <SPAN=20 style=3D"FONT-SIZE: 11pt; mso-ansi-language: DE"=20 lang=3DDE>Ballentine.<o:p></o:p> <P=20 style=3D"TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; mso-hyphenate: none; = tab-stops: -.5in"=20 class=3DMsoNormal>Quinlan, P., van der=20 Maas, H., Jansen, B., Booij, O., = Rendell, M. (2007). =20 Re-thinking stages of cognitive development: An appraisal = of=20 connectionist models of the balance scale task.<SPAN=20 style=3D"mso-spacerun: yes"> <I=20 style=3D"mso-bidi-font-style: normal">Cognition, <I=20 style=3D"mso-bidi-font-style: normal">103, 413-459. <SPAN=20 style=3D"FONT-SIZE: 11pt">Robbins, S. E. (2002). Semantics, experience = and=20 time. <I=20 style=3D"mso-bidi-font-style: normal">Cognitive Systems Research, 3, 301-335. = <o:p></o:p> <SPAN=20 style=3D"FONT-SIZE: 11pt">Robbins, S. E. (2004). On time, memory and = dynamic=20 form. <I=20 style=3D"mso-bidi-font-style: normal">Consciousness and Cognition, = <I=20 style=3D"mso-bidi-font-style: normal">13, 762-788.<SPAN=20 style=3D"mso-spacerun: yes"> <o:p></o:p> <SPAN=20 style=3D"FONT-SIZE: 11pt">Robbins, S. E. (2006a). On the possibility of = direct=20 memory. In New = Developments in=20 Consciousness Research, <SPAN=20 style=3D"mso-bidi-font-style: italic">ed.,V. W. = <SPAN=20 class=3DSpellE>Fallio. Nova Science, 1-64. <o:p></o:p> Robbins, S. E. = (2006b). <SPAN=20 class=3DSpellE>Bergson and the holographic theory of mind.<SPAN=20 style=3D"mso-spacerun: yes"> <I=20 style=3D"mso-bidi-font-style: normal">Phenomenology and the Cognitive=20 Sciences, 5,=20 365-394.<o:p></o:p> Robbins, S. E. (2007). = Time, form=20 and the limits of qualia. <I=20 style=3D"mso-bidi-font-style: normal">Journal of Mind and Behavior, = <I=20 style=3D"mso-bidi-font-style: normal">28, = 19-43.<o:p></o:p> <SPAN=20 class=3DSpellE><SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: = 12.0pt">Savelsbergh<SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">, G. J. P., = Whiting, H.T.,=20 & Bootsma, R. J. (1991).<SPAN=20 style=3D"mso-spacerun: yes"> Grasping <SPAN=20 class=3DSpellE>tau. = <I=20 style=3D"mso-bidi-font-style: normal">Journal of Experimental = Psychology: Human=20 Perception and Performance, 17,=20 315-322.<o:p></o:p> <P=20 style=3D"TEXT-ALIGN: justify; TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; = mso-hyphenate: none; tab-stops: -.5in"=20 class=3DMsoNormal><SPAN=20 style=3D"LETTER-SPACING: -0.15pt; FONT-SIZE: = 11pt">Turvey<SPAN=20 style=3D"LETTER-SPACING: -0.15pt; FONT-SIZE: 11pt">, M., & <SPAN=20 class=3DSpellE>Carello, C. (1995). Dynamic touch.<SPAN=20 style=3D"mso-spacerun: yes"> In <I=20 style=3D"mso-bidi-font-style: normal">Perception of Space and = Motion, eds., W.=20 Epstein & S. Rogers. = Academic=20 Press. = <o:p></o:p> <SPAN=20 style=3D"FONT-FAMILY: 'Times New Roman'">Wigner<SPAN=20 style=3D"FONT-FAMILY: 'Times New Roman'">, E. P. (1970).<SPAN=20 style=3D"mso-spacerun: yes"> Symmetries and = Reflections.<SPAN=20 style=3D"mso-spacerun: yes"> MIT = Press.<o:p></o:p> <SPAN=20 style=3D"FONT-FAMILY: 'Times New Roman'">Woit<SPAN=20 style=3D"FONT-FAMILY: 'Times New Roman'">, P. (2006).<SPAN=20 style=3D"mso-spacerun: yes"> <I=20 style=3D"mso-bidi-font-style: normal">Not Even Wrong: The Failure of = String Theory=20 and the Search for Unity in Physical Law.<SPAN=20 style=3D"mso-spacerun: yes"> Basic = Books.<o:p></o:p> <P=20 style=3D"TEXT-ALIGN: justify; TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; = mso-hyphenate: none; tab-stops: -.5in"=20 class=3DMsoNormal><SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">Woodward, J. = (2000).=20 Explanation and invariance in the special sciences.<SPAN=20 style=3D"mso-spacerun: yes"> British Journal for the = Philosophy of=20 Science, 51, = 197-214.<SPAN=20 style=3D"mso-spacerun: yes"> <o:p></o:p> <P=20 style=3D"TEXT-ALIGN: justify; TEXT-INDENT: -0.5in; MARGIN-LEFT: 0.5in; = mso-hyphenate: none; tab-stops: -.5in"=20 class=3DMsoNormal><SPAN=20 style=3D"FONT-SIZE: 11pt; mso-bidi-font-size: 12.0pt">Woodward, J. = (2001). Law and=20 explanation in biology: Invariance is the kind of stability that = matters.<SPAN=20 style=3D"mso-spacerun: yes"> Philosophy of Science, = <I=20 style=3D"mso-bidi-font-style: normal">68, = 1-20.<o:p></o:p> . </DIV></BODY></HTML>