TNP 5. Distributed Representation

The Network Phenomenon: Empiricism and the New Connectionism
Stephen Downes, 1990
(The whole document in MS-Word)


TNP Part IV Previous Post


V. Distributed Representation


A. A First Glance at Distributed Representation

Above, when discussing the "Jets and Sharks" example, I mentioned that the hidden units represent individual people. This representation occurs not in virtue of any property or characteristic of the unit in question, but rather, it occurs in virtue of the connections between the unit in question and other units throughout the network. Another way of saying this is to say that the representation of a given individual is "distributed" across a number of units.

The concept of distributed representation is, first, a completely novel concept, and second, central to the replies to many of the objections which may be raised against empiricism and associationism. For that reason, it is best described clearly and it is best developed in contrast with traditional theories of representation.

B. Representation and the Imagery Debate

Traditional systems of representation are linguistic. What I mean by that is that individuals and properties of individuals are represented by symbols. More complex representations are obtained by combining these symbols. The content of the larger representations is determined exclusively by the contents of the individual symbols and the manner in which they are put together. For example, one such representation may be "Fa&Ga". The meaning of the sentence (that is [a bit controversially], the content) is determined by the meanings of "Fa", "Ga", and the logical connective "&".

There are two defining features to such a theory of representation. [18] First, they are governed by a combinatorial semantics. That means there are no global properties of a representation which are over and above an aggregate of the properties of its atomic components. Second, they are structure sensitive. By that, what I mean is that sentences in the representation may be manipulated strictly according to their form, and without regard to their content.

There are several advantages to traditional representations. First, they are exact. We can state, in clear and precise notation, exactly the content of a given representation. Second, the system is flexible. The symbols "F", "G" and "a" are abstracts and can stand for any properties or individuals. Thus, one small set of rules is applicable to a large number of distinct representations. That explains how we reason with and form new representations.

We can illustrate this latter advantage with an example. Human beings learn a finite number of words. However, since the rules are applicable to any set of words, they can be used to construct an infinite number of sentences. And it appears that human beings have the capacity to construct an infinite number of sentences. Now consider a system of representation which depends strictly on content. Then we would need one rule for each sentence. In order to construct an infinite number of sentences, we would need an infinite number of rules. Since it is implausible that we have an infinite number of rules at our disposal, then, we need rules which contain abstract terms and which apply t0 any number of instances. [19]

The linguistic theory of representation is often contrasted with what is called the "picture theory" of representation. Picture theories have in common the essential idea that our representations can in some cases consist of mental images or mental pictures. It appears that we manipulate these pictures according to the rules which govern our actual perceptions of similar events. For example, Sheppard's experiment involving the rotation of a mental image suggest that there is a correlation between the time taken to complete such a task and the angle of rotation - just as though we had the object in our head and had to rotate it physically. [20] See figure 7.



Sheppard's experiments are inconclusive, however. Pylyshyn [21] cautions that there is no reason to believe that laws which govern physical objects are the same as those which govern representations of physical objects (he calls the tendency to suppose that they do the "objective pull"). It could be the case, he argues that Sheppard's time-trial results are the result of a repeated series of mathematical calculations. It may be the case that we need to repeat the calculation for each degree of rotation. Thus, a mathematical representation could equally well explain Sheppard's time trial results.

I agree with Pylyshyn that the time trial results are inconclusive. So we must look for other reasons in order to determine whether we would prefer to employ a lingustic or non-linguistic theory of representation.


C. Cognitive Penetrability

Pylyshyn argues that in computational systems there are different levels of description. Essentially, there is the cognitive (or software) level and the physical (or hardware) level. [22] Aspects of the software level, he argues quite reasonably, will be determined by the hardware. The evidence in favour of a language-based theory of representation is that there are some aspect of linguistic performance which cannot be changed by thought alone. Therefore, they must be built into the hardware. Further study is needed to distinguish these essential hardware constraints from accidental hardware constraints, but this need not concern us. [23] The core of the theory is that there are cognitively impenetrable features of representation, and that these features are linguistic.

These features are called the "functional architecture". When I assert that these features are linguistic, I do not mean that they are encoded in a language. Rather, what I mean is that the architecture is designed to be a formal, or principled architecture. It is structure-sensitive and abstract. Perhaps the most widely known functional architecture is Chomsky's "universal grammar" which contains "certain basic properties of the mental representations and rule systems that generate and relate them." [24] Similarly, Fodor proposes a "Language of Thought" which has as its functional architecture a "primitive basis", which includes an innate vocabulary, and from which all representations may be constructed.

The best argument in favour of the functional architecture theory (in my opinion) is the following. In order to have higher-level cognitive functions, it is necessary to have the capacity to describe various phenomena at suitable degrees of abstraction. With respect to cognitive phenomena themselves, these degrees of abstraction are captured by 'folk psychological' terminology - the language of "beliefs", "intentions", "knowledge" and the like (as opposed to descriptions of neural states or some other low level description). The adequacy of folk psychology is easily demonstrated [26] and it is likely that humans employ similarly abstract descriptions in order to use language, do mathematics, etc.

It is important at this time to distinguish between describing a process in abstract terms and regulating a process by abstract rules. There is no doubt that abstractions are useful in description. But, as argued above, connectionist systems which are not governed by abstract rules can nonetheless generalize, and hence, describe in abstract terms. What Chomsky, Fodor and Pylyshyn are arguing is that these abstract processes govern human cognition. This may be a mistake. The tendency to say that rules which may be used to describe process are also those which govern them may b what Johnson-Laird calls the "Symbolic Fallacy".

What is needed, if it is to be argued that abstract rules govern mental representations, is inescapable proof that they do. Pylyshyn opts for a cognitive approach; he attempts to determine those aspects of cognition which cannot be changed by thought alone. The other approach which could be pursued is the biological approach: study brains and see what makes them work. Although I do not believe that psychology begins and ends with the slicing of human brains, I nonetheless favour the biological approach in this instance. For it is arguable that nothing is cognitively impenetrable.

Let me turn directly to the attack, then. If the physical construction of the brain can be affected cognitively, then even if rules are hard-wired, they can be cognitively penetrable. If we allow that "cognitive phenomena" can include experience, then there is substantial experimental support which shows that the construction of the brain can be affected cognitively. There is no reason not to call a perceptual experience a cognitive phenomena. For otherwise, Pylyshyn's argument begs the question, since it is easy, and evidently circular, to oppose empiricism is experience is not one of the cognitive phenomena permitted to exist in the brain.

A series of experiments [has] shown that the physical constitution of the brain is changed by experience and especially by experience in early age. Huber and Weisel showed that a series of rather gruesome stitchings and injuries to cats' eyes changes the pattern of neural connectivity in the visual cortex. [27] Similar phenomena have been observed in humans born with eye disorders. Even after a disorder, such as crossed eyes, has been surgically repaired, the impairment in visual processing continues. Therefore, it is arguable that experience can change the physical constitution of the brain. There is no a priori reason to argue that experience might not also change some high-level processing capacities as well.

Let me suggest further that there is no linguistic aspect of cognition which cannot be changed by thoughts and beliefs. For example, one paradigmatic aspect of linguistic behaviour is that the rules governing the manipulation of a symbol ought to be truth-preserving. Thus, no rule can produce an internally self-contradictory representation. However, my understanding of religious belief leads me to believe that many religious beliefs are inherently self-contradictory. For example, some people really believe that God is all-powerful and yet cannot do some things. If it is possible even to entertain such beliefs, then it is possible at least to entertain the thought that the principle of non-contradiction could be suspended. Therefore, the principle does not act as an all-encompassing constraint on cognition. It is therefore not built into the human brain; it is learned.

Wittgenstein recognized that many supposedly unchangeable aspects of cognition are not, in fact, unchangeable. In On Certainty, Wittgenstein examines those facts and rules which constitute the "foundation" or "framework" of cognition. he concludes, first, that there are always exceptions to be found to those rules, and second, that these rules change ovr time. He employs the analogy of a "riverbed" to describe how these rules, while foundational, may shift over time.

If the rules which govern representations are leaned, and not innate, then it does not follow that it is necessary that these rules ought to follow any given a priori structure. And if this does not follow, then it further does not follow that rules ought to follow a priori rules of linguistic behaviour. Since the rules which govern representations are learned, then, t follows that there is no a priori reason to suppose that human representations are linguistic in form. They could be, but there is no reason to suppose that they must be. The question, then, of what sort of rules representations will follow will not be determined by an analysis of representations in the search of governing a priori principles. We must look at actual representations to see what they are made of and how they work. In a word, we must look at the brain.


D. two Short Objections To Fodor

before looking at brains, I would like in this section [to] propose two short objections to Fodor's thesis that language use is governed by a set of rules which is combinatorial and structure-sensitive.

The first objection is that it would be odd if the meaning of sentences was determined strictly by the meanings of their atomic components. The reason why it would be odd is that we do not determine the meaning of words according to the meanings of their atomic components, namely, letters. It is in fact arguable that even the pronunciation of words is not determined strictly according to the pronunciation of individual letters (here I am reminded of Orwell's "ghoti", an alternative spelling for "fish"). Further, just as the pronunciation of individual letters is affected by the word as a whole (the letter "b" is pronounced differently in "but" and "bought" - the shape of the mouth differs in each case), so may also the meaning of a word be affected by the sentence as a whole and the context in which it is uttered. [28] See figure 8.



The second objection is that a strict application of grammatical rules results in absurd sentences. Therefore, something other than grammatical rules governs the construction of sentences. Rice [29] points out that the transformation from "John loves Mary" to "Mary is loved by John" ought to follow only from the transitivity of "loves". However, the transformation from "John loves pizza" to "Pizza is loved by John" is, to say the least, odd. In addition, grammatical rules, if they are genuinely structure-sensitive, ought to be recursive. Thus, the rule which allows us to construct the sentence "The door the boy opened is green" from "The door is green" ought to allow us to construct "The door the boy the girl hated opened is green" and the even more absurd "The door the boy the girl the bot bit hated opened is green" and so on ad infinitum ad absurdum (or whatever).


E. Brains and Distributed Representations

The human brain is composed of a set of interconnected neurons. Therefore, in order to determine whether or not the brain, without the assistance of higher-level rules, can construct representations, it is useful to determine, first, whether or not representations can be constructed in systems comprised solely of interconnected neurons, and second, whether or not they are in fact constructed in brains in that manner.

At least at some levels, there is substantial evidence that representations both can be and are stored non-linguistically. Sensory processing systems, routed firs through the thalamus and then into the cerebral cortex, produce not sentences and words, but rather, representational fields (much like Quine's 'quality spaces' [30]) which correspond to the structure of the input sensory modality. For example, the neurons in level IV of the visual cortex are arranged in a single sheet. The relative positions of the neurons on this sheet correspond to the relative positions of the input neurons in the retina. Variations in the number of cells producing output from the retina produce corresponding variations in the arrangement of the cells in V-IV. Similarly, the neurons in the cortex connected to input cells in the ear are arranged according to frequency. In fact, the cells of most of the cerebral cortex may be conceived to be arranged into a "map" of the human body. [31] See figure 9.



The connections between neurons in the human visual system are just the same as those in the connectionist networks described above. There are essentially two types of connections: excitatory connections, which tend to connect one layer of cells with the next, as for example the connections between retinal cells and ganglia are excitatory; and inhibitory connections, which tend to extend horizontally at a given level, as for example the connections between the horizontal cells and the bi-polar cells in the retina are inhibitory. [32] See figure 10. Thus, the structure of cells processing human vision parallels the structure of IAC networks described above.



It is, however, arguable that while lower-level cognitive processes such as vision can occur in non-linguistic systems, higher-level cognitive processes, and specifically those in which we entertain beliefs, knowledge, and the like, must be linguistically based. [33] In order to show that this is wrong, it is necessary to show that higher-level cognitive processes may be carried out by an exclusively lower-level structure. The lower level structure which I believe accomplishes this is distributed representation.

I think that the best way to show that these higher-level functions can be done by connectionist systems is to illustrate those arguments which conclude that they cannot be done and to show how those arguments are misconstrued. In the process of responding to these objections, I will describe in more detail the idea of distributed representation.


F. Distributed Representations: Objections and Responses

There are two types of objections to distributed representations.

The first of these is outlined by Katz, though Fodor provides a version of it. The idea is that if representations depend only on connected sets of units or neurons, then it is not possible to sort out two distinct representations or two distinct kinds of representations. Katz writes, "Given that two ideas are associated, each with a certain strength of association, we cannot decide whether one has the same meaning as the other, whether they are different in meaning... etc." For example, "ham" and "eggs" are strongly associated, yet we cannot tell whether this is a similarity in meaning or not. [34]

Fodor's argument is similar. Suppose each unit stands for a sentence fragment, for example, "John -" is connected to " - is going to the store." Suppose further that a person at the same time entertains a connection between "Mary -" and " - is going to school". In such a situation, we are unable to determine which connections, such as between "John -" and " - is going to the store" represents a sentence and which, for example "John -" and "Mary -" represents some other association, for example, between brother and sister. [35]

In order to respond to this objection, it is necessary to show that there can be different types of connections. Thus Katz suggests, for example, that there ought to be meaning connections, similarity connections, and the like (I am here using "connection" synonymously with "associations", which is not exactly right, but close enough: think of an association as a set of connections). If there are distinct types of connections, then there is a distinguishable subset of connections, say, meaning-connections or similarity-connections, which exist necessarily in order to distinguish sentences from, say, similarities. But the set of sentence-connections would be just the functional architecture or primitive basis described above.

There is, however, another way to respond to this objection. First, I would like to suggest that sentence-construction and internal representation are different sorts of activities. Sentence-formation is a behavioural or output process, while representation is a cognitive process. the only way to argue against this suggestion is to argue that representations are irreducibly linguistic. Since this is exactly the issue of debate, it will not do to stipulate that representations must be irreducibly linguistic. So it is permissible for me to suggest, at least as a hypothesis, that the two functions are separate.

Now let me consider how representations such as "John -", "Ham", etc. are stored in connectionist systems. [36] It is not the case that they are stored as sentence fragments, as Fodor misleadingly suggests. Rather, a representation of an individual thing, such as "ham" (ignoring such questions as whether we are talking about one ham or ham in general) is a set of connections between one unit at a hidden layer and a number of other units at a different layer. In other words, a representation is a pattern of connectivity. These patterns may be represented as "vectors" of connections between the hidden unit and a matrix of other units at another layer. Therefore, the representation of "ham" is a set fo active and non-active units connected to a given unit and the activation of the representation is the activation of the appropriate neurons in that pattern. It is often convenient to line up the matrix of units and display the vector which corresponds to the hidden units as a series of 1s and 0s according to whether the units are off and on. So the vector for "Ham" could be represented as "10010010010...10010".

Now we can distinguish between several types of connections. two representations, that is, two units at the hidden level, are "similar" if and only if their vectors of connections are similar. In turn, two representations, that is again, two units at the hidden layer, are in some other way associated is they are both units which constitute part of a vector of a higher layer unit. This is a different sort of association, although the mechanism which produce[s] this distinct type of association is exactly the same as the former. A third type of connection exists between units at a given layer, and that is if the units are clustered with each other via inhibitory connections to form a competitive pool.

Therefore, there can be three types of association which can be defined in a connectionist system: relations of similarity, which correspond to similarities of vector; relations of category which correspond to competitive pools at a given layer [37]; an conceptual relations, which correspond to two different units being a part of the same vector. It should be evident that there can be many types of conceptual relations, in fact, one conceptual relation for each vector for each unit at the next layer. One type of conceptual relation is membership in the same sentence. How this process occurs will be described below. See figure 11.



The second sort of criticism of distributed representation is that if representations are distributed, then two representations of the same thing, say "Ham", might be different from each other (Fodor: "no two people ever are in the same intentional stance" [38]). Suppose then we have two different pattern of connectivity, each of which we'll say stands for "Fa" (of course, it doesn't stand for, or correspond to, the sentence per se, but that's the way Fodor puts it so we'll leave it like that). According to Fodor, one of these must be the representation, while the other is only an approximation of the representation. The problem lies in determining which of these two patterns of connectivity actually stands for "Fa". [39]

In response, one might wonder why there ought to be one and only one meaning or representation of "Fa". Fodor considers this solution to be "the kind of yucky solution they're crazy about in AI". ad hominem aside, it is far from common-sense to believe, as Fodor believes, that there can be one and only one representation of "John Lennon is a better lyricist than Paul McCartney". Further, connectionist systems provide an understanding of how there could be indeterminate representations. This provides a flexibility which language-based systems cannot provide, a flexibility which is essential in our everyday lives.

Let us suppose that a given unit stands for "Fa" (a bit incorrect, but let's suppose). Then this unit may be activated by the vector "11001100". Yet (and this is easily proven on connectionist systems) even a partially complete or incorrect vector will activate the unit in question. For example, it is easily shown that the unit will be activated by "11001101". Fodor's objection consists in the objection that there is no one vector that is the vector that ought to represent "Fa". But that is like attacking a critic of Platonic forms on the ground that there is no means of determining which of two hand drawn triangles is the triangle.

This is an important concept, and I wish to linger just a moment, for it will surface again. Any given concept, and indeed any given representation, needs not correspond to a particular vector, but rather, may consist of a set of vectors. And further, this set of vectors may not have precise boundaries, for what counts as an instance of a concept may vary according to what other concepts are contained in the system. If, for example, there are three colour concepts, then the colour concept "red" may correspond to a wide set of vectors. On the other hand, if there are eight colour concepts, then the set of vectors which correspond to "red" may be much narrower.

In summary, first, I argue that representational structures are learned; they are not innate. Second, I argue that they are distributed, and not symbolic. And third, I argue that concepts are fuzzy, and not precise. The third is what we would expect were the second true, and the second is what we would expect if the first were true. But my third claim can be empirically tested. It is possible to examine actual human concepts, categories and the like in order to determine whether or not they are vague or fixed. If, as I suspect a rigorous empirical examination will show, they are in fact vague, then, first, I have a confirming instance for my own theory, and second, the language-of-thought theory has a serious difficulty with which it must contend.


TNP Part VI Next Post



[18] See Fodor and Pylyshyn, "Connections and Cognitive Architecture: A Critical Analysis", in Steven Pinker and Jacques Mehler, Connections and Symbols, pp. 12-13.

[19] Pylyshyn's "Dial 911 in the event of a fire" example contains much the same argument. There are mny ways to recognize a fire, and many ways to dial 911. It is implausible that we have a rule for each possible case. Thus, we have a general rule which covers these sorts of situations.

[20] Sheppard's experiments and others are summarized in Kosslyn (et.al) "On the Demystification of Mental Imagery" in Ned Block (editor), Imagery. See also Stephen Kosslyn, Ghosts in the Mind's Machine.

[21] in Block, Imagery.

[22] Zenon Pylyshyn, Computation and Cognition. Pylyshyn does not discuss what I call the data level. Others, for example McCorduck, suggest that we ought to contemplate a further "knowledge" level. Pamela McCorduck, "Artificial Intelligence: An Apercu", in Stephen Graubard (editor), The Artificial Intelligence Debate, p. 75.

[23] For example, one accidental feature of the hardware might be the material that it is built from. Theorists who assert that only humans have the appropriate hardware are called identity theorists. See U.T. Place, "Is Consciousness a Brain Process?", and J.C.C. Smart, "Sensations and Brain Processes", both in V.C. Chappell (editor), The Philosophy of Mind (1962), pp. 101-109 and 160-172 respectively.

[24] Noam Chomsky, "Rules and Representations", Behavioral and Brain Sciences 3 (1980), p. 10.

[25] Jerry Fodor, Language of Thought.

[26] See Jerry Fodor, Psychosemantics, ch. 1.

[27] See James Kalat, Biological Psychology (1988), p. 192.

[28] Philip Johnson-Laird, The Computer and the Mind, p. 290, cites A.M. Liberman (et.al.), "Perception of the Speech Code", Psychological Review 74 (1967) to make the point.

[29] Sally Rice, PhD Thesis, cited Jeff Elman, "Representation in Connectionist Models", Connectionism Conference, Simon Fraser University, 1990.

[30] W.V.O. Quine, Word and Object.

[31] All this is surveyed in P.S. Churchland, neurophilosophy and James Kalat, Biological Psychology.

[32] Kalat, Biological Psychology, p.185.

[33] Eg. Jerry Fodor and Zenon Pylyshyn, "Connectionism and Cognitive Architecture: A Critical Analysis" in S. Pinker and J. Mahler, Connections and Symbols. Also, in Anderson and Bower, Human Associative Memory: A Brief Edition (1980), p. 65.

[34] Jerrold Katz, The Philosophy of Language, ch. 5.

[35] Fodor and Pylyshyn, "Connectionism and Cognitive Architecture", in Connections and Symbols, p. 18.

[36] This account is drawn from Rumelhart and MacClelland, Parallel Distributed Processing, Vol. 2, Ch. 17, and from Hinton and Anderson, Parallel Models of Associative Memory.

[37] Categories are also defined by similarities. This is discussed below.

[38] Fodor, Psychosemantics, p. 57.

[39] Fodor, Psychosemantics.

Read More

TNP 4. Connectionism

The Network Phenomenon: Empiricism and the New Connectionism
Stephen Downes, 1990
(The whole document in MS-Word)


TNP Part III Previous Post


IV. Connectionism

A. Basic Connectionist Structures

A connectionist system consists of a set of neurons, or "units", and a set of connections between those units. The units may be activated or inactivated, Most systems employ simple on-off activations, although other systems allow for degrees of activation. The motivation for this basic structure is biological. Connectionist systems emulate human brains, and human brains consist of interconnected neurons which may be activated (spiking) or inactivated.

The idea is that a unit "i", if activated, sends signals or "output" via connections to other units. Other units, in turn, send signals to other units, including unit "i". These signals comprise part of the "input" to "i". It is also possible to provide input to "i" via some external mechanism, in which case the input is referred to as "external input". For any given unit "i", the state of activation of "i" at time t depends on its external and internal input at time t and the state of activation at time t-1. Input can be excitatory or inhibitory. If it is excitatory, then the unit tends to become activated, while if it is inhibitory, then the unit tends to become inactivated.

Connections between units may similarly be excitatory or inhibitory. If a connection between two units "i" and "j" is excitatory, then if the output from "i: is excitatory, then the input to "j" will be excitatory, and if the output from "i" is inhibitory, then the input to "j" will be inhibitory. If the connection between "i" and "j" is inhibitory, then excitatory output will produce inhibitory input, and inhibitory output will produce excitatory input. See figure 1.



Typically, units in a given network are arranged into visible and hidden layers. The visible layers are in turn divided into input and output units. The idea is that the hidden units are sandwiched between the input and the output layers. The input units are activated by external input, and these in turn activate appropriately connected hidden units. These hidden units in turn activate output units. Thus, one might say that input stimulus produces output response. See figure 2.



b. Pools

The designation of one or another set of the visible units into input or output units is to some degree arbitrary. It is often more useful to think of visible units as being divided into "pools" such that any given pool or set of pools may be a set of input or output units depending on circumstances. The idea is that units in a given pool may be connected to each other and to units at higher or lower levels, but not to units in other pools at the same level.

The basic idea is derived from Feldman. [12] Suppose we have two units "i" and "j", each with a single input and a single output. Excitatory input will activate the unit and it will in turn send excitatory output. Suppose now that each unit is connected to the other such that if "i" is activated, it will tend to inhibit "j", and if "j" is activated, it will tend to inhibit "i". Then over time, whichever unit re3ceives more input activation will tend toward maximum activation, while the other will tend toward minimum activation (that is, maximum inhibition). A network like this is called a "Winner-take all (WTA)" network. See figure 3.



If you do this with two or more units, you have a pool. An example of this sort of structure is McClelland and Rumelhart's "IAC (Interactive Activation and Competition)" network. [13] This is an interesting network because it shows how networks can categorize and generalize.

McClelland and Rumelhart use as an example a network called "Jets and Sharks". Each unit at the visible level stands for some property of a person, for example, his age (in20s, in30s, in40s), his occupation (burglar, pusher, etc.), his name, his education, and so on. Each of these sets of properties (age, occupation, name, education) constitutes a single pool. Units in a given pool are connected to other units in the pool and to units in a second, hidden layer of units. the connections have been predefined such that connections between members of a given pool are inhibitory and connections between the visible and the hidden layers of units are excitatory. See figure 4.



The idea is that each unit at the hidden layer stands for an individual person. What characterizes that person (that is, the knowledge stored about the person) is not some property of the unit in the hidden layer, but rather, the set of connections between that unit and the units at the visible layer. Take, for example, some hidden unit "i". This unit contains no information about the person. However, it is connected to the units at the visible layer standing for Jake, burglar, in20s, high school, Jet (the gang name), etc.

Suppose now that the network could learn the connections just described. Then it would have leaned a system of categorization. Each of the pools constitutes a distinct category. The fact that it i a category is established by the inhibitory connections between all and only units of a given pool and by the fact that one, and only one, unit in a given pool is connected to any given unit at the hidden layer. What makes, say, in20s, in30s, and in40s a single category is first the fact that they have something in common - they are all associated with some person - and second that they are mutually exclusive. The activation of one inhibits the rest.

The IAC network can perform a number of associative and inductive tasks. For example, suppose we activated "burglar", "in20s" and "high school". Then, because of the excitatory connections to a given hidden unit, that unit would become activated. In turn, the hidden unit would send excitatory output via an excitatory connection to the visible unit representing "Jake". Thus, by input to a set of features, an individual's name may be recognized. What is interesting is that the name may be activated even if the description is incomplete or incorrect. [14] The reliability of such a conclusion drawn in such circumstances varies according to the scale of the missing or incorrect information and according to other connections in the network.

Such a network can also generlaize. For example, suppose the unit for "Jet" were activated. This unit is connected to a number of hidden units, and these will be activated. Each of the hidden units will send excitatory output to units in the other visible pools. Several units in each pool may be excited. However, since the connections between the units of a given [pool] are inhibitory, then only the unit with the greatest excitatory input will be activated; the rest will be inhibited, Thus, upon the activation of "Jets", a set of units, one in each pool, will also become activated. This set of units is a "stereotypical" picture of the Jets. For example, activating "Jets" may result in the activation of "in20s", "pusher", "high school", etc. One might say that these features constitute a definition of the category "Jets" even though no individual Jet has all and only those stereotypical features.

A similar sort of network performs well in visual recognition or multiple constraint tasks. One could input observed features of a person at a distance and the output could be that person's name. As with the previous example, the nature and reliability of the conclusion will vary according to circumstances. For example, a distinctive walk could very quickly aid in the determination of a given person's name, but if the system has information about two such people with the same distinctive walk, then this determination will not be so quick and so sure.


C. Learning in Connectionist Networks

In my mind, the real advantage of connectionism does not lie in the features just described, for those features could be realized by a system with enough predefined rules. Rather, the advantage is that such a system can learn its own connective structure. A network learns by adjusting connection w3eights between different units. There are several ways of doing this, and this accounts for one of the major differences between types of connectionist systems.

The simplest sort of system employs the Hebb rule. According to the Hebb rule, if two units are simultaneously activated, then the connection between thm should b strengthened. Similarly, the connection between two units should be strengthened if the two units are simultaneously inhibited. If the two units are at a given time at different states of activation (one is excited, the other inhibited) then the connection is weakened. The major problem (in my mind) with the Hebb rule is that it doesn't work in networks with more than two layers. This is a substantial weakness, since as Minsky and Papert point out, a two-layer network cannot distinguish between, say, exclusive and inclusive disjunctions. [15]

Most contempoary systems use a version of the "delta-learning rule". In such a system it is important to distinguish between input and output units. Input units are excited and consequent output noted (if there are more than two layers and no connections are yet established there will be no output at first). The output obtained is compared to the desired output and an error is computed. From this error a correction can be calculated (take the error and apply a linear or non-linear function to it. This produces a curve, the slope of which will be zero at the point of minimum error. There are various strategies for going 'down' the curve.) This correction is then propagated back through the network and the correction is distributed across the connections which contributed to the error. This process is called back-propagation.

Such a system depends on a teacher. This sometimes seems to pose a problem for artificial intelligence theorists who would rather the machine learned completely on its own. [16] However, experience can teach. The concept of "lessons of nature" dates back at least to the Scholastic philosophy of the middle ages and is a central thesis of empiricism. One can imagine how some particular output (a response or behaviour) might require correction because it causes, or fails to prevent, pain.

One of the dificulties encountered in back=propagation systems is the problem of the "local minima". What happens when you calculate the error curve for a number of variables is that you might not get one single location where the curv is zero; you might get several. One such point may be still an error, but since the slope i zero, there is no means for the system to correct itself, since the degree of correction is usually a function of the slope of the curve. What you want to do is to "shake" the system so that it reaches the lowest minimum. See figures 5 and 6.






This is accomplished in two stages. First, each unit is considered to have two possible states of activation, namely, activated and inactivated. Each unit has another state, which is its probability of activation. Input from other units or from external units affects the probability of activation and not the state of activation itself. Then, in the second stage, the probability that a given nit will be activated is represented by the function:



where E stands for the "energy" of the system and "T" stands for the "temperature" of the system. [17]

The reason why the terms "energy" and "temperature" are employed is that the equation above is borrowed directly from physics. Essentially, the higher the temperature, the more random the activation or inactivation of a given unit will be. As it turns out, if a system is started at a high temperature and then, as processing continues, the temperature is lowered, the system is much less likely to settle into a local minimum. This process is called "annealing" and is exactly analagous to the physical process (used to produce stable crystalline formations) of the same name.

There are several useful features of this process which I won't detail, however, I will mention that these is an equation such that the energy (and hence, error) of any given connection can be determined. Hence, energy minimization (and hence, error correction) can be accomplished at a local level, with no regard to the global properties of a system. This means that "higher level" knowledge is not required for error correction.


TNP Part V Next Post




[12] Feldman, J.A. and D.H. Ballard, "Connectionist Models and Their Properties", Cognitive Science 6 (1982), pp. 205-254; cited in Alvin Goldman, "Epistemology and the New Connectionism", in N. Garver and P. Hare, Naturalism and Rationality (1986), p. 84.

[13] Parallel Distributed Processing I, p. 28.

[14] This is called content addressability and differs from traditional systems. See James Anderson and Geoffrey Hinton, "Models of Information Processing in the Brain", in Anderson and Hinton, editors, Parallel Models of Associative Memory, p.11.

[15] Marvin Minsky and Seymour Papert, Perceptrons.

[16] Eg., Hinton, who proposed that learning could be accomplished without training if the system had a built-in coherence requirement. Geoffrey Hinton, "The Social Construction of Objective Reality in a neural Network". Connectionism conference, Simon Fraser University, 1990.

[17] This equation is specific to "Boltzmann" machine versions; there are other equations that accomplish the same thing. These are described in Parallel Distributed Processing I, ch. 6 and 7.

Read More

TNP 3. Three Objections to Empiricism

The Network Phenomenon: Empiricism and the New Connectionism
Stephen Downes, 1990
(The whole document in MS-Word)


TNP Part II Previous Post


III. Three Objections to Empiricism

A. Objections to Associationism

There are three essential objections to empiricism. The first objection is that associative principles are not sufficiently powerful to explain human cognition. The second is that there is no means to distinguish input from other cognitive phenomena. And the third is that associative inferences can never be justified. I will discuss each in more detail and show what I need to prove in order to meet the objection.

The first objection always has the following form: "Human beings can know or do X, no associative system can know or do X, therefore, humans use something other than associative principles."

A paradigm example of such an objection mentions humans' use of abstractions. Consider the following example. [5] Suppose we had to determine whether a string of letters is a well formed formula (wff) in a language L. Language L is a "mirror image" language; only strings the two halves of which are mirror images are wffs in L. In order to distinguish wffs in L from non-wffs in L, it is necessary to employ an abstract term. Since it is impossible for an associationist principle to employ an abstract term, then the principle we employ to distinguish wffs and non-wffs cannot be an associaionist principle.

Another example may be found in Leibniz. [6] While we need experience to suggest to us universal eneral principle principles, we cannot derive these principles from experience. For experence consists entirely of particulars, and no set of particulars is ever sufficient for the derivation of a universal. Therefore, we must emply some means other than experience in order to derive universal general principles.

These examples could be multiplied, but they give the general idea. Since the argument is valid, then the only means of answering such objections is to deny either (a) that human beings can know oe do X, or (b) that associationist principles are insufficient for X. In general, I take the following approach. If the claim is that we know X, then I deny 9a). If the claim is that we do X, then I deny (b). Classical scepticism is all that I need to deny (a). The real challenge lies in the denial of (b).


B. Theory-Laden Perceptions

The second objection to empiricism reaches the conclusion that, since there is no means of distinguishing perception from other aspects of cognition, it follows that we cannot say that other aspects of cognition have their origin in perception.

The premise is well supported, For example, Quine argues that we cannot distinguish between the analytic (for example, formal rules of inference) and the synthetic (empirical content). [7] A similar point is made by Hanson. According to Hanson, what we see is affected by what we believe, that is, all our experiences are "theory-laden".

Since the premise of this argument is well supported, the only means of responding to this argument is to show that the conclusion that there are no pure perceptions does not follow from the premise. There are two ways of stating the premise. I will consider each in turn.

The first way is to state the premise is to say that perceptual terms are theory-laden. [9] The second is that perceptions themselves are theory-laden. The first formulation I embrace, since all aspects of language are theory-laden. Language depends to a large degree on rules and categories, and these, I believe, are theories. However, it does not follow from this premise that there are no pure perceptions, since perceptions are distinct from descriptions of them.

The second way of stating the premise is to the effect that perceptions themselves are theory-laden. I can agree that at some level, perceptions are theory-laden. This is a natural and expected consequence of the theory of learning which I will propose below. If it is true that perceptions are always theory-laden, then the conclusion, that there are no pure perceptions, follows. So, in order to show that this conclusion does not follow, I will need to show, first, that there is some level of perception that is not theory-laden, and second, there was some point in time at which no perception is theory-laden.

This will not be an easy task. It is arguable, for example, that even if there are pure perceptions, they cannot be used unless combined with some theoretical structure or another. [10] It is also arguable that in order to perceive objects in three dimensions, some higher-level built-in constraints are required. [11] Therefore, in order to meet this objection, not only is it necessary to show that perceptions are pure at some level at some time, it is also necessary to show that these perceptions are sufficient for all other cognitive activity.


C. Justification

The whole idea behind justification is that of distinguishing between right and wrong (correct and incorrect, justified or unjustified) inferences, and the third objection to empiricism is that it cannot distinguish between right and wrong inferences.

The ground for this objection is that associationism does not distinguish between truth-preserving infrencs and other sorts of inferences. For example, suppose we adopt a causal theory of cognition (following, say, Armstrong and Goldman). Some causal interactions, for example, the triggering of relays in a computer, are truth-preserving. Others, for example, a bat striking a ball, are not.

What is needed, the objection continues, is a formal representation of the sort of causal interactions which are truth-preserving, in order to distinguish fropm those which are not. This representation must exist at a level over and above mere physical instantiation. An example of the sort of principles that we require is the set of rules of logical inference.

In response to this argument, I will argue, first, that the distinction between right and wrong inferences is sufficiently drawn by the concept of relevant similarity, and second, that associationist systems allow only inferences which preserve relevant similarity. Therefore, associationist systems, by the fact that they are associationist systems, provide sufficient justification for their conclusions.

An objection to this conclusion may state that there is a rather large difference between truth and similarity, and thus one cannot equate a mechanism which preserves similarity with a mechanism which preserves truth. Therefore, associationist systems do not provide sufficient justification for their conclusions.

The reason why I believe that justification must be defined in terms of similarity are complex. They will be discussed below. At this point, however, may I say that, if it is indeed the case that justification can be defined in terms of similarity, then the third objecyion can no longer be sustained.


TNP Part IV Next Post


[5] T.G. Bever, J.A. Fodor, M. Garrett, "A Formal Limit of Associationism," from Verbal Behaviour and General Behavour Theory, T.R. Dixon and D.L. Horton, editors. Prentice-Jall, 1968.

[6] New Essays Concerning Human Understanding.

[7] W.V.O. Quine, "Two Dogmas of Empiricism", From a Logical Point of View.

[8] N.R. Hanson, Patterns of Discovery.

[9] Eg., Paul Churchland, "Two Kinds of Evidential Bias". I have only a manuscript of this.

[10] Lawrence Bonjour. The Structure of Empirical Knowledge.

[11] Marr, Vision. See also Phillip Johnson-Laird, The Computer and the Mind.

Read More

TNP 2. Empiricism

The Network Phenomenon: Empiricism and the New Connectionism
Stephen Downes, 1990
(The whole document in MS-Word)


TNP Part I Previous Post

II. Empiricism

A. What I Mean By Empiricism

When I speak of "empiricism" I wish to make it clear that I am not discussing logical positivism or other contemporary theories which have been described as empirical. Rather, what I mean has a much closer affinity to the philosophies of David Hume and John Stuart Mill. To employ Hume's terminology, what I wish to assert is that all ideas are copies of impressions. Modern terminology demands a more precise definition.

We may identify three distinct levels of human cognition. [1] The first, or lowest level is the "hardware" level, that is, the physical structure in which cognition occurs. The second ;evel is the "software" level, that is, a set of rules or procedures which govern cognitive processes. Third, there is the "data" level, which contains the contents of cognition, for example, mental representations.

Empiricists assert that all content at the data level has its origin in experience. What this means is that all content must have been, at some time or another, input through one or another of the senses. The contemporary dispute between empiricists and other philosophers concersn the origin of the software level. Empiricists believe that these rules or processes are learned, while other philosophers believe that they are innate or otherwise directly intuited.

Insofar as we are talking about formal rules, for example, the rules of logical inference, grammar, or mathematics, then I am in agreement with the empiricists. However, I believe that these rules belong to the data level. I think that they describe and do not govern. The rules or principles which actually govern cognition are of a different type: they are associative, not formal, principles. In this way, I believe my thesis differs significantly from contemporary or positivist forms of empiricism.

Like Hume, I believe that human cognition is governed by human nature. Thus, I believe that the associative principles which govern human cognition are a part of, or instantiated in, human nature. Therefore, according to the sort of empiricism I am proposing, instead of there being three levels of cognition, there are only two lvels: the hardware level, which contains the associative principles which govern cognition, and the data level, which contains the content of cognition including descriptive 'formal' principles.

Why am I the sort of empiricist I say I am? First, I do not believe that human cognition is governed by formal rules. Otherwise, we would never be able to break these rules, and there is substantial evidence that we can. So I believe we are not governed by formal rules, and further, I believe that these rules cannot be innate.

And second, even were we governed, innately or otherwise, by formal rules, I would argue that we should not be. For formal rules are abstractions, and while abstractions are useful, they are insufficient to respond to sufficiently complex phenomena. As Wittghenstein [2] points out, we can always find an exception to a rule, and we need to be able to respond effectively even in the case of an exception.


B. Association, Rules and Categories

Before the advent of logical positivism, empiricists such as Mill and Mach argued that general principles, such as formal rules or laws of nature, are summaries of previously experienced phenomena. [3] In my opinion, this view of rules is correct. Rules, as Wittgenstein argued, describe, and they do not prescribe.

Generalized descriptions such as rules or laws of nature may be derived employing associative principles. Simply put, the idea is that, when we observe a sequence of similar events in which two things go together, we generalize and say those things generally go togethr.

There is in my mind a close link between rules and categorizations. When we place two things into a category, we are saying that those two things are similar in some way. We use categories in order to generalize. If most members of a given category are associated with something, say, some sort of behaviour, we tend to associate all members of the category with that behaviour.

The concept of "similarity" is central to empiricism, for all association and categorization depends on similarity. By "similarity", I wish to emphasize, I do not mean identity of a set of observational predicates. In my opinion, similarity is a pre-linguistic concept. below, I will say that two things are similar if they have sufficiently overlapping vectors.

The entire principle of associationism may be defined by the following paradigm example: "dpgs are similar to cats, dogs are associated with fur, therefore, cats are associated with fur." While this may seem to be a very weak principle, once it is recognized that the 'cat', 'dog' and 'fur' in the example can be anything, for example, '1101', '1100', and '0001' respectively, then we can see that this is a very powerful principle.

TNP Part III Next Post


[1] I use the term "levels" in much the same manner as Pylyshyn, Computation and Cognition.

[2] On Certainty. "A rule is shewn by its exception."

[3] John Stuart Mill, A System of Logic, and Ernst Mach, The Analysis of Sensations.

[4] Ludwig Wittgenstein, Philosophical Investigations.

Read More

TNP 1. Introduction

The Network Phenomenon: Empiricism and the New Connectionism
Stephen Downes, 1990
(The whole document in MS-Word)



1. Introduction
2. Empiricism
3. The Objections to Empiricism
4. Connectionism
5. Distributed Representations
6. The Problems of Perception
7. Associationism: Cognitive Structures
8. Associationism: Inferential Structures
9. Connectionism and Justification
10. Summary
11. Projects and Investigations
TNP: 20 Years On

I. Introduction

I wish to argue in this paper that the new connectionism provides a vindication for classical empiricism. By "empiricism" I mean the philosophy that all knowledge has its origin in experience, that is, that there is no innate or otherwise intuited knowledge. My argument is that connectionism provides a computational framework within which traditional objections to empiricism may be met.

The structure of this paper is as follows. I will begin with a description of what I mean by empiricism. Then I will set out a series of objections to the philosophy I describe. In order to meet those objections, I will first describe connectionism, then outline some objections particular to connectionism, and then finally respond to each of the objections mentioned. Finally, I will discuss some further avenues of investigation.

I would like to caution the reader that this is to a large degree a survey paper. While arguments are sketched in order to demonstrate the plausibility of the thesis being proposed, I do not claim to have completely solved all the problems and to have met all the objections. In addition, the reader will note that many of the arguments given are only sketches and do not consider a number of important yet intricate points. The reason for this is that length was, believe it or not, a limiting factor in this presentation.

TNP Part II Next Post

Read More

School Choice

Responding to Joanne Jacobs:

Without school choice, Ty’Sheoma Bethea will stay in her second-rate school

What does she think would happen if 'school choice' should suddenly appear? That this one person - and no other - would go to the first-rate school? No, of course not - but then, would everyone go to the first rate school? No, that wouldn't work either - there aren't enough spaces, and creating them would ruin the first-rate school.

The reasoning, of course, is that choice would create competition, which would magically make underpaid and underfunded schools suddenly become better. As good as the first-rate schools, even - because, otherwise, the logic simply doesn't work.

In fact, it doesn't work at all. The idea if school choice being the answer to someone stuck in a 'second-rate' school is a farce. You won't make all schools first-rate, and you won't get nearly all of the students into the first rate school.

The only way school choice makes sense is in supporting the *type* of education that is more appropriate for people (this, though, doesn't fly with the standards crowd because it allows that people have different learning styles, different needs, different interests, and that these could be served by the school board).

The fact is, "school choice" - at least how it is being used here - is code for "private". And these days, the people supporting privatization bear the onus of proof. The privatization crowd has basically wrecked the economy and the parts of the school system they touch - like, say, Edison schools - often end up as a wreck as well.

There is such a thing as genuine choice. I wrote about it here: http://www.downes.ca/post/44259 But it has nothing to do with privatization, and everything to do with quality education. So it's probably not of interest here.

Read More

The Monkeysphere Ideology

Goodness knows, I don't want to cite Cracked as my source.

My brother liked Cracked. I was always a Mad reader. He also liked Pepsi and CFGO in Ottawa. Myself, I was always a devotee of Coke and listened to CFRA. The originals.

But sometimes you follow the leads wherever they take you. But let me digress first.

Unlike most people, I did not lose any money in the economic crash. At least, no money that I know of yet - I may have something in some pension account somewhere. But I don't have investments, retirement accounts, or any of that sort of thing.

"I won't get to retire," I always said when people asked me. "Whatever retirement money I could ever save, they would figure out some way to steal it." And so they have, and now with the Dow passing 7000 and continuing its downward plunge, it feels like I'm watching the fall of a civilization.

I watched The Day After on YouTube today. You can watch the entire length of the controversial 1983 made-for-TV movie. The story centers around the survivors of a nuclear attack. The few that made it endured chaos, disease and hardship. Their entire way of life disappeared in jyst a few minutes.

I have always wondered why people go on with their daily lives in cities on the brink of disaster. How the residents of Pompeii, for example, were cooking bread and weaving cloth right up to the time of the fateful eruption. How villages continued as normal up to the very minute of Genghis Khan's golden horde.

Now I know: what else can you do? The disasters cut a swatch through society, everything changes, and then you try to make do with whatever you have left.

So what does this have to do with Cracked?

Well - it's this. Great societies, they endure. Their fabric withstands the blows of fate and fortune and there is enough in their people to carry on after. To carry on in an altered, reformed, fundamentally different state, perhaps, but to carry on.

But what gives you that capacity is not typically your technology or your wealth or your dominions - all of which are characteristically wiped out in a crash. No it is your character, your capacity not simply to carry on, but to have a reason to carry on, to rebuild what you have lost.

Now let's look at the Cracked article, which suggests that each of us has a limit of about 150 people we can know and understand and relate to. The theory is based on Dunbar's number, and Cracked calls it - with more than a little alacrity - the 'monkeysphere'. The article, which was written in 2005, is making the rounds again.

In our complex society, writes Cracked, "Most of us do not have room in our Monkeysphere for our friendly neighborhood sanitation worker. So, we don't think of him as a person. We think of him as The Thing That Makes The Trash Go Away."

So far, this is fine. We have limits to our capacity. We are monkey brains. We all know that. But the writer takes it a step further. " We are hard-wired to have a drastic double standard for the people inside our Monkeysphere versus the 99.999% of the world's population who are on the outside."

My fiend Bob Armstrong used to say, when we worked on the Gauntlet together, that the importance of a story to the media was inversely proportional to the distance from us and proportional to the amount of blood shed and the whiteness of their skin. An observation, not a thesis, and one that remains true today, at least in western media.

But being Cracked, the thesis is pressed one fatal step further: "The problem is that eventually, the needs of you or those within your Monkeysphere will require screwing someone outside it (even if that need is just venting some tension and anger via exaggerated insults). This is why most of us wouldn't dream of stealing money from the pocket of the old lady next door, but don't mind stealing cable, adding a shady exemption on our tax return, or quietly celebrating when they forget to charge us for something at the restaurant."

Except... that's not true.

Oh, wait a minute. It's true for some people, at least. "There is a reason why all of the really phat-ass nations with the biggest SUV's with the shiniest 22-inch rims all have some kind of representative democracy (where you vote for people to do the governing for you) and all of them are, to some degree, capitalist (where people actually get to buy property and keep some of what they earn). "

And this is what I have always known about what would be the fate of my putative retirement savings plan.

And what we are going through now is the logical consequence of thinking like monkeys. If we can't even get though a day without yelling at people on the road, stealing money from old ladies, or cheating on our taxes, cable bills or restaurant cheques, then any hope we have of building a modern technological society is probably doomed. They're too fragile. They require a high degree of intelligent behaviour on the parts of their citizens.

And we have spend the last few decades fostering, nay celebrating, the ethos of the monkeysphere. Believing that if each of us looked out solely and entirely for our own interests (and that it wasn't cheating unless you got caught and convicted). A nation of Conrad Blacks, looking at us smugly, derisively, snarling at our inability to understand the realities of our times.

And even as our society heads toward the precipice, life continues on as usual. Television channels continue to play Jerry Springer and Dr. Phil. The newspapers continue to publish stories about the needs of business and retirement savings. Our society continues to slide - and, one thinks, it will not cease to slide until people get the point.

The point is that the monkeyspehere ethos that has been informing our society over the last three decades or so is fundamentally wrong.

Our failure lies not in the fact that we cannot know and understand more than 150 people. That's just a fact of physiology. Rather, our failure lies in how we characterize the remaining 99.99 percent of humanity: as though they were automatons.

This is the fundamental error of our times. It is the error that allows us to characterize entire societies as 'ragheads', the culture that allows is to say "it's not personal, it's business" as we evict someone from their home or cheat them out of their life's savings, the ethos that allows us in the western world to build a society based on consumption and ownership of more and more even as starvation and disease wrack the remainder of the world.

This is what allows us to treat politics and warfare as games, that allows people like Rush Limbaugh to say he hopes Obama's plan will fail, that allows us to treat education as though it were economics, able to be sceptical about reform but not really caring, because those kids, aren't people, beacuse success has nothing to do with lives, everything to do with test scores.

This fallacy persists. The failure to understand just what has gone wrong with our society continues in our government halls, where our own ministers are sacrificing humanities and the arts for business - “[s]cholarships granted by the Social Sciences and Humanities Research Council will be focused on business-related degrees.”

And I read this headline in he New York Times: "In tough times, the humanities must justify their worth." Not business, which caused this mess, not media, which propagated the monkeysphere ideology, not accounting, law or political science, which participated by stumbling around each other in their haste to see who could be corrupted most quickly. No - humanities. And arts.

Because the humanities continue to be portrayed as the pastime of the idyll rich: "a traditional liberal arts education is, by definition, not intended to prepare students for a specific vocation. Rather, the critical thinking, civic and historical knowledge and ethical reasoning that the humanities develop have a different purpose: They are prerequisites for personal growth and participation in a free democracy, regardless of career choice."

Our falure is not a failure of business, which performed as intended (at least for those who made off with the wealth). It is a failure of the humanities, a failure of humanity, the study of which has been in notable decline throughout these last few decades, having, if you will, no measurable worth, no valuation, it being nothing more than a pastime and a recreation.

Ironic then that Obama's success in the United States is the very antithesis of that: “He does something academic humanists have not been doing well in recent years,” [Andrew Delbanco] said of a president who invokes Shakespeare and Faulkner, Lincoln and W. E. B. Du Bois. “He makes people feel there is some kind of a common enterprise, that history, with its tragedies and travesties, belongs to all of us, that we have something in common as Americans.”

The case has been made before. Our media, one of the early victims of the rise of corporatism, has transformed us from a society of thinkers and reflectors to a society of passive consumers of slapstick. A society were a Jerry Springer retort is what constitutes a reasoned argument, a society where lies and deception become standard fare in the media, a society in which cardboard caricatures substitute themselves in our awareness for reality.

I find myself asking this a lot, "How can you find that moral?" or "how can you find that ethically defensible?" Business practices that depend on preventing poor people from obtaining an education. World financial systems that require deceptive advertising and child labour in work camps in the third world. A network of luxuries and resources that are based in the systemic looting and deprivation of entire populations.

Andrea and I went out to see The Reader last night, a film that had enough conflict, sex and nudity to catch the attention of the box-office-sensitive critics and Oscar voters. "What would you have done?" asks the illiterate prison camp guard quite reasonably. "There were more people coming. We had no place to put them. What would you have done differently? Should I have not joined the SS?"

What is society, other than law? We are tempted to say that it must be more - that it must be morality, say - but even that is far to shallow a notion. Law and morality are not what make us obey even the little principles that create a society. Law and morality depend, even in themselves, on self-interest, on reward and punishment, on monkey teleology. We know that when law and morality are all that hold us together, things fall apart as soon as the source of order is removed.

Our society is founded, and made possible, though an act of mind: and that act is the capacity to empathize - to see, through reason, the conseuqence of our action on others, and to feel the impact of those consequences in ourselves. We even have bits of monkey brain specifically designed for that purpose. Until we shut them off. Until we deliberately erase their impact, because they have no 'value'.

This is not simply a matter of schooling, not simply a matter of going to college. Perhaps there was a time when we could afford to have a society where education was available only to the elite. It isn't even a matter of preparing students "for professional success, responsible citizenship, and fulfilling lives." It's not a matter of preparing at all.

James Bloom writes, 'When we start telling students, their families and the public who pay for our services: 'Trust us. Don’t ask questions. We know what we’re doing,' instead of encouraging them to ask, 'Why do you what you do?' or 'What’s the point of studying literature and philosophy?,' we’ll deserve to go out of business." But it isn't a matter of being or not being in business.

The economy is just numbers. Education is just facts. Business is just commerce. None of these will offer our society any sort of hope in the current crisis, or the numerous crises that are coming. Yes, the economy is crashing, yes, millions of people are losing their homes and their jobs, yes, we may be only weeks and months away from riots in the streets and civil insurrection - none of this is at the core of our despair.

I finished Hemingway's For Whom The Bell Tolls last night. A story of civil insurrection, of the end of society and the rise of fascism, of casual murder, betrayal, and love. "'Then you will have to fight in your country as we fight here.' 'Yes, we will have to fight.' 'But are there not many fascists in your country?' 'There are many who do not know they are fascists but will find it out when the time comes.'"

In Hemingway's time, as in our own time, society falls, and fascism rises, when the humanity is erased from its citizens. "The soldiers using those weapons are simple brutes, they lack 'all conception of dignity' as Fernando remarked. Anselmo insisted, "We must teach them. We must take away their planes, their automatic weapons, their tanks, their artillery and teach them dignity".

When we live our lives in the monkeysphere, we have no comprehension of any of this. We see glimpses only of the lives of the participants, and mostly, see that they do not see each other as people - as hurting, feeling beings. "Because thou art a miracle of deafness....It is not that thou art stupid. Thou art simply deaf. One who is deaf cannot hear music. Neither can he hear the radio. So he might say, never having heard them, that such things do not exist."

What we need, to survive this crisis and the next, is to get beyond the crass calculations of statistics and value, beyond the idea of "proving your worth", beyond seeing people as caricatures, as cardboads cutouts populating the backdrop of our lives, but of beings worth of consideration, nay, worthy of sacrifice.

This is more than "a common enterprise, that history, with its tragedies and travesties." This is, rather, a way of seeing the world, or as Wittgenstein would say, a way of being, a way of living. Our fundamental bedrock assumption must be, as Kant said, that we treat people as having inherent value in and of themselves. "Act in such a way that you treat humanity both in your own person and in the person of all others, never as a means only but always equally as an end."

I have, from time to time in the past, advocated that educators ought to adhere to something like the Hippocratic oath, a commitment to, above all, do no harm. This ought to be the end of statistical education, the end of the idea that students are not mere caricatures, the end of the idea that educational innovation that satisfies the needs of the many, or the needs of society, or the needs of business or the rich, can be accomplished by the sacrifice of even one person.

And, were a similar standard adopted in our processes of politics and business, it would be the end of government by statistics. The end of the depiction of unemployment as a rate. The end of the accounting of poverty as a percentage. The end of the idea that "it's just business" when we sacrifice a life, and the beginning of the idea that, not only is it morally and legally wrong, it is also fundamentally opposed to our idea of selves as humanity. Inhumane.

So how do we get there?

Ideologically, we have to get beyond the mass. We have to get beyond the idea of seeing ourselves as being nothing more than the corporate entity to which we belong, whether that entity be a business, a religion, a discipline, a nationality. We are each of us members of all of these things, and more, and yet they form only the shallowest part of ourselves.

Yes, though it is empowering and aspirational to be a part of something that 'greater than ourselves', it is key and fundamental to understand that, whatever this thing may be, it is a fiction, an artifice, that we have created in order to more efficiently express our thoughts, feelings and affiliations. The moment we subvert ourselves to the mass, is the moment we can see all other humans as similarly subverted.

Conceptually, we need to begin to think and reason and act in terms of the concrete rather than the abstract. That does not entail the end of abstract reason - far from it - but rather it is to foster in ourselves a clear and precise understanding that the abstract is an artifice, an invention, that we use to facilitate thought and reasoning.

Probably the most evident of the abstract that has become reality, and the form of artifice most often promulgated in our mass media, is that of simple causality, whether that of a war, a depression, a successful education, an election. We hear constantly the idea of some 'leader' or 'great person' (usually from MIT or Princeton or something) having 'done' something, whether it be as mundane as raising money through a football program or 'the inventor of' as though there were no players or society or funding of staff or support or janitorial service that made it possible. Our mass media idolize the famous, and in so doing, relegate the rest of us to being bit players. Props.

This understanding, this way of seeing the world, is wrong, and demonstrably wrong. Place Alexander Graham Bell in the Middle Ages and - guaranteed - he does not invent the telephone. Place Rene Descartes in the Tsarist Russia and the Meditations never sees the light of day. Under slightly different circumstances, Elisha Gray is the inventor of the telephone, or Blaise Pascal the inventor of Cartesian Geometry.

When we magnify the importance of the corporate entity above all else, we hurt society. And when we magnify the importance of the individual actor above all else, we hurt society. The monkeysphere ideology is based on both of those fallacies. Business, media, and the rest of them, are based on that fallacy.

Practically, we must immerse ourselves in our own humanity. We must talk to each other. We must communicate with each other. We must be open about our own lives, and curious about others. We must transcend the limit of the monkeysphere by constructing for ourselves concrete understandings of what it is to be human, to live, to have hopes and dreams, and to die. We must read each others' stories, listen to each others' music, to, above all, communicate.

We may not be able to know, in a personal sense, more than 150 people. But we can know of many many more, and we can know, in a concrete sense, that each of these people live lives of value, and cannot simply be thrown away or discounted, not through any sense of law or ethics, but because that's how it feels to be human.

Which returns us to the monkeyspehere. Which returns us to the fact that most of us would not cheat on our neighbours, steal from the blind, swindle old ladies, and all the rest of it, not because it's against the law, not because it's immoral, but because of the way we feel when we do it. As Hume attempted to explain, our sense of humanity and decency is based on a sense of feeling inside us, a passion, and this passion is in turn born in us through a process of experience and education, the process of living in the world, interacting with others, and understanding them.

In our society today a great many people live without this sense of feeling for others. It's a sad thing, and the result of decades of deliberate desensitization. These are the people who, above all, will not be able to comprehend the economic collapse (or global warming, or resource scarcity, or the rest of it) and, with its onset, will be poorly placed to survive it.

These are the people who, through the decades of the monkeysphere, laughed at us from their SUV, blames poverty on the indigent, and championed the unique acumen and skill of CEOs who, by luck and a similar narrowly focused ethic, managed to steal success and create an empire. These are the people who will be least stable in the coming years, which is why reconciliation - as hard as it may be - will have to serve as a touchstone for our post-crash society.

And the other touchstone will be even more simple and more basic - the preservation and promotion of individual human worth and dignity, for each and every person in society - no exceptions. The understanding that our first response to the crisis will have to be to ensure that everyone remains housed and healthy, nourished and educated. The understanding that acquisition and hoarding are dysfunctional, that the chronically wealthy are, in a certain sense, disabled, and that the wealth of society is the birthright of each and every individual of which it forms a part.

Read More