Tuesday, 24 June 2014

A Gibsonian analysis of linguistic information

This post is based on a talk I just gave at the Finding Common Ground Conference at the University of Connecticut. Please excuse the Power Pointy nature of some sections! You might need to Ctrl+ to see some of the images clearly.  I have made some changes from the original talk content on the basis of very useful feedback I received from other conference attendees.

What is the place of language in ecological psychology? Is language a type of direct perception? Is language comprehension direct perception? Does language have affordances?

In trying to answer these questions I discovered that some things we think of as being perceptual have a lot in common with the conventionality of language and that some language-related behaviours look a lot like perception (as typically construed). I end up suggesting that we move away from talking about 'perception' and 'language' as different types of entities and instead focus on information / behaviour relations in specific tasks.

 In general, ecological language people are happy to talk about language as having affordances. Two papers from the 2010 special issue of ecological psychology on Ecological language even have the word in their titles ("New affordances for language: distributed, dynamical, and dialogical resources" Hodges & Fowler, 2010; "Speech as the perception of affordances" Worgan & Moore, 2010). However, there is a long-standing awareness of the fact that language challenges some important aspects of ecological psychology. As Fowler (1986 p 24) says:

"As to...whether a linguistic message can be said to be perceived...from a direct-realist perspective, direct perception depends on a necessary relation between structure in information media and its distal source. But,...this does not appear to apply to the relation between sign and significance."

What Fowler meant was that because the relation between linguistic structures (e.g., spoken words) and how they structure behaviour is based on convention (not natural laws) they cannot support direct perception. It is clear that language does not fit cleanly into an ecological framework as currently defined. I argue that, rather than vaguely using the vocabulary of ecological psychology to describe language, it's crucial to identify precisely where the similarities and differences are between language and perception. I want to preserve the theoretical rigor of ecological psychology. 

The goal of the following analysis is to devise an extended framework for ecological information that accommodates language without straining or redefining original notions of affordances or direct-perception. 

Comparing language and perception

To begin, let's sum up the basic ecological approach to perception as conceived of by Gibson (1979) and reinforced by Turvey et al (1981). 

1) Affordances are dispositional properties of objects and events that provide opportunities for action to complementary organisms

2) Perception is of affordances

3) Affordances uniquely structure energy (specification)

4) Therefore, coordinating behaviour wrt information is equivalent to coordinating behaviour wrt the affordance property (direct perception)

Illustration of typical perception case:

Illustration of typical language case:

In the perception case, direct perception possible because the task (tree climbing) was about the property of the world that created the information. Behaviour was coordinated WRT the property of the world that caused the information.

In the language example, however, the task is not about the property of the world that created the information. Behaviour is organised according to conventions constrained by culture, linguistic context, etc. If we look closely at the types of behaviour that might arise as a result of receiving the verbal instruction "Turn left at the corner" something interesting emerges. The instructions can have a consequence on our actions - we can initiate a left turn at the corner instead of going straight or right. Or, we might respond, "No, I'll go straight because it's faster" or something similar. Or, we might ignore it altogether (although ignoring such a statement in this kind of conversation is very unusual in our culture). The point is, whatever consequence the information has, we definitely won't be using it to control our left turn. The turn itself will be controlled by good-old-fashioned perceptual information. "Turn left at the corner" can influence action selecting, but it can't be used to control the action. I've intentionally used "action selecting" rather than "action selection", even though "action selection" is the accepted phrase in the literature. "Selection" implies that the information is evaluated by the system, which then selects the most appropriate action. This implication should be avoided. "Selecting" implies that it is the information, itself, that selects the action by interacting with an embodied, extended cognitive system. More on this later. 

Turvey et al (1981) said:

"Our strategy, as proponents of Gibson's ecological approach, is to argue for a conception of natural law that allows meaningful relations between organism and environment to hold. Further, we constrain our use of the term 'perception' (and thus, of course, 'direct perception') to relations governed by such laws."

The view of language that I propose reserves these terms for cases where these requirements are clearly met. Does this mean that language is outside the remit of ecological psychology? No, because for Gibson, the primary thing is information. And, Gibson (1979 p 255) offers a useful suggestion for how to progress our understanding of things like language.

"Perhaps if [so-called higher mental processes] are reconsidered in relation to ecological perceiving they will begin to sort themselves out in a new and reasonable way that fits with the evidence."

We've identified a key difference between perception and (at least some) language-related behaviours. Bearing this in mind, let's now go back to the beginning - which, for Gibson is that the environment is informative - and build up from there.

Re-analysis of information

1) Objects and events structure energy

2) These structures are lawfully related to the properties in the world that cause them

3) Information is any structure in energy that has a consequence for behaviour

4) Not all behaviours are organised wrt the property of the world that caused the information

5) Nervous systems are for linking information and behaviour

Point 4 simply states the difference identified between the perception and language examples above. The first 4 statements broaden the field of information. The 5th statement claims that linking information and behaviour is precisely what we are built to do - variations in nervous system type and complexity can be understood in terms of their consequences for linking information and behaviour. 

Perception / action is a particular type of information / behaviour relation. As I intend to show, calling some types of things 'perception' and other types of things 'language' obscures  important characteristics that more clearly define information types. This means that the key objective is to accurately characterise relationships between information and behaviour. What factors influence the learning of various information / behaviour relations? What type of behaviours is a type of information capable of supporting? What types of information is a particular organism capable of detecting or coordinating its behaviour WRT?

As a starting point, I've identified the following factors as important.

Evolutionary factors

  • the types of energy an organism is sensitive to
  • biases for an organism to pick up certain types of information
  • predetermined or strongly constrained relations (e.g., reflexes / instincts)
  • sophistication of nervous system - ability to learn new information / behaviour relations, to use information of varying degrees of reliability, stability
Learning factors

  • must the organism learn to detect the structure?
  • must the organism learn to organise its beahviour WRT the structure?
  • is the behaviour organised WRT the property causing the structure or convention?
  • how stable / available / reliable is the structure?
  • is the information used in action selecting or in the continuous control of action?
These factors can be used in the form of questions to ask about the types of general information / behaviour relations an organism demonstrates (evolutionary factors) and the types of information / behaviour relations that are specific to a particular task. I'll illustrate how this works using just the learning factors.

The example of a human infant learning to walk is a paradigmatic perception / action task. All typical perception / action tasks should have similar answers to the ones above.

 Notice that the language-related behaviour of postural entrainment looks just like the infant learning to walk from this perspective. Thus, the information used to control postural entrainment during conversation is the same type as the information used to control locomotion.

Not all language-related behaviours are cases of perception / action, though. As we saw earlier, responding to the instructions "Turn left at the corner" involves organising behaviour according to convention. This type of information influences action selecting, but it cannot be used in the continuous control of action.

Interestingly, producing the phrase "Turn left at the corner" looks a lot more like a perception / action task. Whether the task involves action selecting or control appears to depend on the level of analysis. At the phoneme level, it seems to be about action-selecting (this is based on a talk at the conference by Brian Gick which seemed to show that, at least, some of these units were produced with only feedforward control). I don't know a whole lot about this area, but it's worth identifying that the level of analysis matters here.

This example focuses on bees who observe the Waggle Dance and then search for food in a particular direction as a result. This is the first example where the answer to the first two questions is 'no' and it shows the work that evolution can do in constraining conventional information. Still, this conventional information is only involved in action selecting, not in online control. 

Other behaviours that do not require learning to coordinate behaviour wrt the structure include reflexes and (some) instincts. Reflexes and instincts differ in that information for reflexes is for action selecting and information for instincts is for action control

These questions can also be applied to conditional learning, as in this example where bees were trained to associate a particular colour with food. What's nice about this framework is that it provides a common vocabulary for looking at the whole range of animal behaviours.

Quick re-cap and implications

Perceptual information, ecologically defined, does not encompass the full range of information / behaviour relations. In particular, it excludes cases where behaviour is organised according to convention, as with some language-related behaviours, animal communication systems, and conditional learning. Some language-related behaviours (e.g., language-related entrainment behaviours) are controlled by perceptual information, as defined by Gibson & Turvey et al. And, some non-language-related animal behaviours (e.g., behaviours resulting from operant conditioning) are selected by conventional information. This means that a clear perception / language divide is misleading. Information / behaviour relations are more clearly organised according to the factors that influence the learning of those relations, as in the examples above. 

A major implication of this analysis is that many aspects of language do not "have affordances." Affordances are dispositional properties of objects or events that structure energy and provide opportunities for action to organisms with complementary effectivities. Their relevance to ecological psychology is only to establish that there are properties in the world that can support the online control of action. However, not all information / behaviour relations are about the continuous control of action - many are about action-selecting. The theoretical work that affordances need to do is not required in these cases. Nothing is gained by trying to broaden the definition of affordances to accommodate these examples.

A second implication is that many aspects of language do not qualify as direct perception. Direct perception is specifically when behaviour is organised wrt the property of the world causing the information because this entails that organising behaviour wrt the information is equivalent to organising behaviour wrt the property of interest. People have begun using direct perception as shorthand for "not mediated by representations," but it is only one way of being non-representational. Acknowledging that some aspects of language are not examples of direct perception does not necessitate a representational account of language. Nothing is gained by trying to loosen the definition of direct perception to accommodate these examples.

Finally, while this analysis draws out many differences between language-related behaviours and traditional perception / action examples, these are really just two regions on a single field of ecological information. For all learned information / behaviour relations, from walking to talking, an organism must learn to organise its behaviour wrt information in the environment. To re-iterate the point made earlier, this is what nervous systems are for. 


  1. I'm undecided on this, but feel still quite fond of the slightly altered definition of direct perception as an assertion of theoretical stance, rather than characterising a 'type' of perception.

    Surely the 'direct' of 'direct perception' is only helpful if it's counter to the 'indirect' of standard cognitive science models (i.e. internal mediation)? If an extra criterion has to be met (re: natural law shaping info) to warrant being 'direct', then within our own paradigm we would have to define what 'not direct' means in cases where this isn't met, and why we don't want to call it 'indirect'.

    That seems much more messy to me. Fowler says something similar (1996, I think) by saying that if it's as simple a distinction as the historical origin of the structure, it doesn't make it a fundamentally different type of perception.

    While there might be obvious differences in behaviour precipitated by conventional/natural information because of the structure, it makes more sense to talk about it at the level of your taxonomy. Characterising the activities in terms of what ends up sounding like categorically distinct 'types' of perception is scaffolding our thinking in the wrong way - something which the flexibility of your taxonomy avoids.

  2. A few rather disjointed thoughts on this:

    I don't think it's worth loosing the rigour of the current definition of direct perception. That term is doing a lot of work in explaining how prospective control is possible without cognitive enrichment. If it just becomes shorthand for "non-representational" then this argument will still need to be made, and someone will have to draw out the distinction that I've made here, except from the perception/action side.

    I agree that it's important to avoid creating a false dichotomy, and I have very little interest in using the term "direct perception" at all. I'm moving away from talking about linguistic versus perceptual information for the same reason - task specificity is the way forward. The part of the story that isn't told as clearly in this post is the first person story that shows how the processes involved in learning to use information are the same for all types of tasks.

    If we take a theoretical stance that both typical perception / action behaviours and language-related behaviours are non-representational, then we need a good story for how that works. One already exists for perception / action in the form of traditionally defined direct perception. This account cannot be extended to many language related behaviours because of the difference identified early in the post. It simply won't work. Behaviours based on conventional information need their own non-representational explanation. Extending the term direct perception to cover these behaviours may temporarily obscure this, but it leave these ideas open to being skewered by critics.

  3. I think the definition itself as a distinction is fine, I just don't think they should get to keep the term 'direct perception'. I strongly suspect the label was chosen purposefully to oppose 'inference', and that's our shtick for the whole of cognition. In the event of ecological psych taking off, the etymological context will be lost and the term will only be confusing.

    This is mostly me just having a strop, because I like the concept of 'directness' to explain the entire approach. Also since there's plenty of non-linguistic conventional perceptual information that precipitate our actions, it's not even an account for the whole perception-action domain. They are being greedy!

    1. The problem is this: direct perception has a specific meaning. It is the kind of perception in which you have unmediated access to psychologically relevant properties of the world by virtue of the relationship between the proximal stimulation of your sensory systems and the property in question. Specifically, it is the kind of perception you can do when the information you are detecting is both specific to the property in question and is about the property in question. This kind of perception is a) a specific kind of interaction with information within Sabrina's taxonomy and b) only possible if the property creates the information by virtue of a natural law.

      So organising your behaviour with respect to the conventional meaning of some information is not direct perception. Only organising your behaviour with respect to information that is 'about' the thing that created it is an example of direct perception. Using affordances to organise your behaviour entails direct perception; as I'm reading Sabrina, using language to precipitate your behavour is not.

      All of that is the 3rd person analysis. The 1st person experience is 'detect information, organise behaviour with respect to that behaviour, have that work or not work, iterate until success'. In the affordances case, you will succeed if you behave as if the information is about the property that created it. In the speech case (for example) you will fail if you behave that way, but you will succeed if you behave as if the information is about the conventional meaning.

      So (as I make this up as I go): perhaps the 1st/3rd person thing is important here. From the 1st person perspective there is no actual difference in what you do (as Sabrina noted). From the 3rd person perspective there are important differences; the directness varies.

      Thoughts? That idea evolved some as I typed :)

    2. Andrew, William James would point out that that your description does not much resemble first-person experience. The first person experience is "catching the ball". ;- )

    3. Agnes,
      I don't think the distinction between "direct" and "indirect perception" is historic in the way you suggest. It long pre-dates the "cognitive revolution". Indirect perception happens when, for example, by virtue of my sensitivity to light, I point out objects in a picture, or enjoy a show on television. It is "indirect", because the things I am responding to are not there, and I don't know how to describe it other than with words like seeing or vision... so there you go... "indirect perception".

      A big problem happened in perception when renaissance artists and philosopher-scientists decided to use that obviously mediated mode of seeing as a theoretical model to explain all seeing. Similar ideas had certainly been floated before, but never in so formalized and ubiquitous a manner. Bad stuff. Wrong direction. Centuries of confusion. But they had discovered a lot about indirect perception, and we shouldn't pretend that they were wrong, or that that process doesn't happen. We just need to figure out how to convince people that all sorts of different interesting stuff happens during direct perception.

      As Andrew points out, we, as modern ecological psychologists, can see all sorts of fascinating thing from our third person perspective that was overlooked for a long time.

      Note, also George Orwell's warning about the air of mystery that surrounds some Latin-based terms. If we could have this entire conversation with words like "see" or "hear" instead of "perception", that would help a lot.

    4. Fair point, Eric :) It's the 1st person analysis, perhaps.

  4. Hi Sabrina -

    If you've been following the comments on Andrew's last post you know that I'm mostly on-board with your general approach. However, I have a couple of issues with your presentation.

    In the case of language, the scenario comprises an object emitting aural energy (AKA a speaker) and an organism sensing that energy (AKA a hearer). The hearer detects the information in the energy. I simplify by assuming that the speaker utters a simple imperative so that if the information and the context constitute an index that can be used to select a previously learned context-dependent behavioral disposition, that disposition is executed.

    Once a "behavioral disposition" has been selected for execution, presumably the resulting task must be completed if possible. (Isn't that the significance of mandatory effecting of an affordance?) If so, then failure to complete the task must be the consequence of an external event, not of a whimsical decision by the organism. TIn which case "the information ... definitely won't be used it to control our left turn" isn't correct. It won't be used to exercise continuous control, but it does control the hearer's behavior in the coarse sense that the intended task gets completed.

    My impression is that "information" is being used in two different ways. One is the (ecological?) information in the aural energy's structure, and that information is used to select a disposition to execute. And there is information that is carried by grammatical structure. That duality presupposes a certain way of processing spoken language. At least in the case of simple utterances, it seems likely that the ecological information in the query's aural energy is used directly to identify and execute a learned disposition. To the extent that there's any "meaning" in the query, it's merely the intended response.

    Generalizing this idea, one can think of any relatively simple utterance as being used by a speaker to cause a hearer to respond in a certain way, ie, to effect execution of a specific behavioral disposition. From that perspective, the ecological information in the aural energy of an utterance is about the speaker's intent. Intent can be viewed as a behavioral disposition, which being a neural structure is "a property of the world". Then although it's true that the ecological information in spoken language isn't (in general) about the speaker's vocal chords (the entity I assume that in the case of speech is being referred to by "property of the world that created the information"?), that's irrelevant. The speaker's intent causes specific aural energy to be created; if the hearer understands the linguistic content, then the intended disposition is executed. Paraphrasing the non-language case in these terms, the object's dispositional "intent" causes creation of an affordance which if "understood" by the organism causes it to execute the "intended" complementary disposition.

    The particular utterance - hence, aural energy pattern - that effects a particular disposition is, of course, a social convention both across languages and often even within a language. But the process of creating that causal relationship is based on properties of the world via a process Donald Davidson calls "triangulation". In the simplest case, two creatures observe an object or event in the world and agree on a sound to make when reference to that entity is intended. Davidson elaborates this idea in the essay "Three Varieties of Knowledge", concluding that successful communication requires that participants each "know the others mind" in that they mostly use the common language according to rules dictated by the developmental history that relates language to the world. In that sense language is a product of natural law.

    Like Andrew says in his last comment, I'm winging it a bit here, but hopefully this provides food for thought.

    1. I just checked out this post from Eric C's blog that Andrew referenced in his last post but that I had overlooked, and it seems right on point with that last paragraph in my comment:


      Eric's post addresses the possibility that minds are in some sense "visible" and thus can be - again, in some sense - "perceived. I'm suggesting that Davidson's Varieties of Knowledge essay describes how that might work and in essence argues that the relationship between a speaker's mental state and the information that a hearer perceives in the speakers verbal utterances is more law-like than it may seem.

    2. Charles,
      If I understand you correctly, then that is indeed the necessary assertion. This also flows from Bob Shaw's work in Eco-psych (which I mention in a few papers), as well the work of others in allied approaches. To see someone's mind is (by Holt and, I assert, Gibson's analysis) to see what aspects of the world their behavior is a function of. There are, no doubt, difficulties in making such observations, but there are not irresolvable mysteries of the type Descartes, and those who follow his lead, would have you believe.

      Some of those "law like" relations will be very stable over very long periods of time, others will be transient. This connects directly with Andrew and Sabrina's approach in which the organism switches between different types of task-specific devices.

  5. Hi Sabrina,

    It was great seeing your talk at UCONN and it was helpful to listen to an oral presentation of these ideas, which supplemented your previous blog post. Many people talked very highly of your presentation and ideas, so thanks for making the trip.

    My question is whether or not this framework can, or should, be applied within an empirical setting. If so, is there a simple example you could provide to aid in understanding what an empirical level of analysis would look like? It is really enjoyable to see how this work is evolving!

    Best Regards,
    Drew Abney

    1. Hi Drew,
      Great to hear from you here!
      Yes, this framework is meant to be applied in an empirical setting. Broadly speaking, Andrew and I are in favour of a 4 step empirical approach to understanding a given behaviour (see here for a paper on this http://journal.frontiersin.org/Journal/10.3389/fpsyg.2013.00058/abstract).

      The evolutionary and learning based factors identified above are meant to complement this 4 step process. Many of the factors can be translated into a research question about a particular information / behaviour relation. The clearest examples of these concern whether an organism needs to learn to detect the information and whether the organism needs to learn to organise its behaviour wrt the information. Some of the evolutionary factors are also testable - e.g., whether there are any biases to pick up particular types of information, the complexity of information / behaviour relations an organism is capable of learning.

      Basically, the 4 step process is the place to look to identify what information an organism uses to accomplish a particular task. And, the factors above are useful for identifying why that particular information / behaviour relation exists (is it learned, is it constrained by evolution, was this information variable the only sufficiently stable one, etc).

  6. Sorry to be barging in late, but... Yes, this!

    "rather than vaguely using the vocabulary of ecological psychology to describe language, it's crucial to identify precisely where the similarities and differences are between language and perception. I want to preserve the theoretical rigor of ecological psychology. "

    Some of the rest seemed like a call for more human ethology, which I am always in favor of.

    Also, be careful with the the learning thing. Everything we are talking about develops, and we don't want to get stuck splitting hairs about what the more narrow term "leaning" means.

    Hopefully I will have a chance to catch up with the rest of the comments soon. In the meantime: Great talk!

    1. Yes, everything develops and there is often not a clear break between learned and unlearned, but it is worth paying attention to and ecological psychologists have missed a lot by ignoring it. This is clear in Karen Adolph's summary of infant stepping behaviour. Stepping begins in the womb, very early in development. It seems like a behaviour that just happens in mammals. And then, somehow, walking happens a little while later. How should we understand this transition?

      First, stepping is only the behaviour half of an information / behaviour relation. Second, stepping doesn't equal walking. But, why doesn't stepping equal walking? Because walking involves coordinating a stepping behaviour wrt particular types of perceptual information. Babies must first learn to detect this perceptual information (they do this during the many months they spend not walking) and then the must learn to coordinate their behaviour wrt it. Paying attention to the learning angle helps us understand what is really driving the development of the target behaviour.

  7. Hi Charles,

    A few things...

    Since I think this analysis shows that conventional information doesn't have affordances, I am not troubled by the whole "compulsory effecting of" issue.

    I don't think intent can be viewed as a behavioural disposition. First, an intention would have to be a definable thing (an object, event, or relation) to have dispositional properties. I could imagine someone making the argument that an intention is an event, but this is doing precisely what I argue against - straining or broadening well-defined terms from ecological psych. I suppose the move is to say that a speech event is the information for an intention? This doesn't change the fact that the link between that intention and the specific form of the information (particular word used, etc) is conventional rather than natural.

    The natural law account of affordances requires behaviour to be organised with respect to the property in the world causing the information. The idea of complementarity is that just knowing about the dispositional property is sufficient to tell you what types of actions it would be possible to effect on the basis of that property. This is why this is a useful theoretical move that is worth preserving.

    With a speech event, the articulatory dynamics creating the sound do not lawfully constrain the types of actions it is possible to effect. Yes, there are constraints (culture, learning context, etc), but they are conventional. There is no theoretical ground to be gained by acting as if these constraints are based on dispositional properties. It doesn't help explain linguistic behaviour.

    To my mind, the more useful thing to notice is the divide between action-controlling and action-selecting information. Affordances are the necessary underpinning to action-controlling information. Conventions are sufficient for action-selecting information.

  8. Since there are some fundamental disagreements here, I've been trying to model nore carefully my view of how both the non-linguistic and linguistic cases work . Here's my take which is based on viewing both cases as using a communication channel. Having selected this model, these questions need to be answered:

    1. What is the purpose ("meaning", if you like) of the communication?

    2. Who/what are the users, ie, the sender and the receiver?

    3. What is the nature of the information to be communicated?

    4. What is the transmission medium for the information?

    5. How is the information applied to the medium (ie, how does the information modulate the medium)?

    Answering several of these questions is straightforward: in both cases, I view the purpose of using the channel to be controlling and/or selecting the behavior of an animal. And therefore, I see the receiver as being a subset of the animal's brain that includes elements of the sensory and motor cortices, in particular neural structures I call "context-dependent behavioral dispositions" (henceforth, simply "dispositions"). In the non-linguistic case the medium is ambient light which is modulated by being reflected off the surface. Since the spatial relationship between surface and animal can be time-varying, the information can be as well. That completes the model in the non-linguistic case.

    The linguistic case is trickier, which I think explains some of our disconnects. I see the linguistic sender as also being a subset of an animal's brain that includes elements of the sensory and motor cortices, including dispositions. However, whereas activating relevant dispositions in the receiving animal requires external sensory stimulation, for execution of relevant dispositions in the sending animal internal stimulation alone (in the vocabulary of psychology, desires, beliefs, needs, wishes, etc) is sufficient. Answering the questions about information and medium is trickier still. For simplicity, I'll focus on the special case in which the information to be sent is a simple imperative, ie, a request for the receiving animal to execute a specific disposition which is identified by a specific linguistic structure - word, clause, or sentence. The association of the disposition to be executed and linguistic structure is assumed to have been learned through the process of triangulation described in my comment of 27 June at 13.41. The linguistic structure is sent from sender to receiver via some medium: sound, writing, light, et al, but which one - and therefore its specific mechanics - is irrelevant since the information is in the linguistic structure, not in the dynamics of its modulation of the medium.

    Presumably the important feature of a lawful relationship between received information and features of the "sender" (surface or neural structure) is reliability: in sufficiently similar contexts, sufficiently similar information will result in similar behavior by the perceiver. While the choice of language is indeed conventional, once a language has been chosen successful communication requires that it's use be subject to certain constraints. And the point of triangulation is that language does have a high degree of reliability because it's learned by a process of interaction among speaker, hearer, and world. Therefore, effective use of a language must be consistent with the way it was learned and hence with characteristics of the world. Of course, there are many reasons why communication may fail. For example, the actual state of the receiver may not be that assumed by the sender. But owing to the nature of learned behavioral dispositions, that has to be the exception rather than the rule. And communication in the non-linguistic case can fail as well. The fielder can stumble or misjudge the ball's flight despite the information being created in accordance with natural law.