Chapter 5 Flashcards

1
Q

What two theories are key to embodied cognitive science?

A

Situatedness (how an environment is sense) and embodiment of the agent (or how the agent is able to interact in the world, which is dependent on the agent’s physical form).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is posthumanism?

A

The result of when the content of information is more important than the physical medium in which is represented; specifically it is when consciousness is considered to be epiphenomenal, and where the human body is considered just a prosthetic.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Describe methodological solipsism

A

The individuation of representational states only in terms of their relation to other representational states. Relations of the states to the external world - the agent’s environment - are not considered; in psychology this is the view that psychological states should be construed without reference to anything beyond the boundary of the individual who has those states

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is the parable of the ant?

A

The theoretical situation of observing an ant on the beach. Initially one might think that the ant is producing complex behaviour, a series of complicated twists and turns that comprise the ant’s route. Classical cognitive scientists tend to explain these type of complexities by invoking complicated representational mechanisms.

Simon, in contrast, noted that the path of the ant might in fact merely result from simple internal processes reacting to complex external forces - the various obstacles the terrain itself provides. The complexity of the ant is really a complexity in the surface of the beach.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What are Shapiro’s key three themes which he argues are taken up by embodied cognitive scientists to varying degrees?

A
  1. Conceptualization: the concepts that an agent requires to interact with its environment depends on the form of the agent’s body. With different bodies comes different understanding/engagements with the world.
  2. Replacement: the idea that we do not need to have a representative world in our heads as it is already accessible in the environment.
  3. Constitution: the body and the world are literal constituents of the cognitive process.This connects to the extended mind hypothesis.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What is the extended mind hypothesis?

A

Is the belief or idea that because cognition can take place in the environment, our mind has therefore ‘leaked’ into the world around us, and the borders between mind and environment are blurred/disappeared - our mind has been extended.
» added onto this notion is the concept behind cognitive scaffolding - if we are externalizing our cognition and using information processing in this manner, than cognition is therefore “our mind” + these external processes

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is forward engineering/synthetic psychology?

A

The manner in which embodied cognitive science explores its theories of cognition - they build simple agents and then observe them in environments of varying complexity, which provides cognitive science with more powerful and usually much simpler theories (induction takes more time than deduction; analysis is harder than invention).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

How is forward engineering different from reverse engineering?

A

This is different from reverse engineering, which involves observing the behaviour of an intact system in order to infer the nature and organization of the system’s internal processes. It usually relies on internal processors to explain behavioural complexity. With forward engineering a system is constructed from a primitive functions of interest, and then the system is observed for any interesting or surprising behaviour. Reverse engineers collect data to build their models, whilst forward engineers build models and rely on them for data.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What is the travelling salesman problem and why is it important?

A

The TSP is a vital optimization problem which involves determining the order in which a salesperson should visit a sequence of cities, stopping at each city only once, such that the shortest total distance is travelled. The problem is important because its solution can be applied to various real-world problems and situations (ie. scheduling tasks, minimizing interference amongst a network of transmitters, data analysis in psychology, warehouse order-picking problems, etc.); as the number of cities involved in the salseman’s tour increases linearly, the computational effort for finding the shortest route increases exponentially = NP-complete problem.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

How do ants solve the travelling salesman problem?

A

By interacting with the world and basic behaviours; as the ant moves it deposits a pheromone trail which decreases in potency over time. The ant on the shortest trail will therefore deposit a trail at the decision points sooner than the ant on the longest one, and will therefore reinforce (with scent) the short route. Over time this intensification increases as more and more ants are attracted to the stronger pheromone trail, and it quickly becomes the selected route.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What is the theory of holism?

A

Where the regularities governing a whole system cannot be easily reduced to a theory that appeals to the properties of the systems part; Gestalt psychologist popularized the notion of “the whole is not merely the sum of its parts”, and Wheeler proposed the notion of the ‘superorganism’, which is comprised of individual colony members but has higher-order properties that cannot be reduced to these component parts.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

How is the construction of wasps nests an example of dynamic environmental control guides?

A

The individual’s behaviour is a program whose actions are under some type of environmental control; there is no need for direct communication amongst the insects because they act on the environment in a manner which signals to the others to act on it in another way.
In the instance of the wasps, the insects build a nest consisting of a lattice of cells, where each cell is essentially a comb created from a hexagonal arrangement of walls. The decisions to add new cells is driven by the perception of existing ones, and two simple rules (theorized by Theraulaz and Bonabeau):
1. if there is a location on the nest in which three walls of a cell already exist, then this acts as a stimulus to prompt the wasp to add another wall there with high probability
2. if only two walls already exist as part of a cell, this also acts as a stimulus, albeit one with less probability

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Define stigmergy

A

The notion that the environment is a stimulus that causes particular work, or behaviour to occur (a circular action, in which the agent changes the environment, which elicits new behaviour, which changes the environment, which now elicits different behaviour, and so on). Stigmergy only works when an agent can sense the world, and when they can physically act on the world.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What does it mean to be embodied to different degrees?

A

Some definitions of embodiment relate to the extent to which an agent can alter its environment; embodiment is grounded in the relationship between a system and its environment. The more a robot can perturb an environment, and be perturbed by it, the more embodied that robot is. In the same argument, a robot that is more embodied than another is more capable of affecting, and being affected by, its environment.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Describe feedback and how it works in embodied cognition.

A

The power of embodied cognitive science emerges from agents that are situated and embodied because these two features allow for a source of non-linearity called feedback. Feedback occurs when information about an action’s effect on the world is used to inform the progress of that action (the circularity of action exists between the parts of a dynamic system).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What is the issue with utilizing feedback as an important component of cognition?

A

It essentially makes computational analysis extraordinarily difficult due to intractability of the mathematics of even a small number of interacting components. Ashby (1956) argued that a number as small as four produced a system which could not be analysed. This is precisely why embodied cognition employs forward engineering as a method of composing theories.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

What is behind the idea of uphill analysis and downhill synthesis?

A

This is terminology used by Braitenberg, and refers to the problem of doing analysis with induction, and the ease of performing inventive measures using deduction (which follows a straightforward path).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

What is umwelten?

A

A term used to describe the “island of senses” produced by the unique way in which an organism is perceptually engaged with its world; because different organisms experience the world in different ways, they can live in the same world but at the same time exist in different umweltens.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

What is the idea of ecological theory of perception?

A

The notion that one cannot separate the characterisitics of an organism from the characteristics of its environment (animal and environment make an inseparable pair).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Describe affordances

A

The affordances of an object are the possibilities for action that a particular object permits a particular agent. Affordances emerge from an integral relationship between an object’s properties and an agent’s abilities to act.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

What is enactive perception?

A

A reaction against the traditional view that perception is constructing internal representations of the external world; rather, it argues that the role of perception is to access information in the world when it is needed. It is a sensorimotor skill.

22
Q

What role does action play in enactive perception?

A
  1. perception is not to be viewed as building internal representations of the world, but instead as controlling action on the world
  2. our perceptual understanding of objects is sensorimotor; that is, we obtain an understanding about the external world that is related to its changes in appearances that would result by changing our position (by acting on an object or by moving to a new position)
  3. perception is to be an intrinsically exploratory process; therefore we do not construct complete visual representations of the world but instead have we have access to properties in the world when needed, and only through action
23
Q

Describe subsumption architecture.

A

It is a set of modules, each of which is a sense-act mechanism; that is, every module can have access to sensed information as well as to actuators. There is no separation of perception from action, but instead, each module is used to control some action on the basis of sensed information. Subsumption architecture arranges modules hierarchically. Lower-level modules provide basic, general-purpose, sense-act functions. HIgher-level modules are more complex, and have more specific sense-act functions which exploit the operations of the lower-level modules (think the ability to run a motor vs. the ability to steer…despite the latter being higher-level, it cannot function without the former). Vertical sense-act modules are the foundation of subsumption architecture.

24
Q

What is an example of subsumption architecture in the human brain?

A

The dorsal and ventral streams of vision - dorsal is meant for action, whilst ventral is meant for object identification. You can damage one ‘module’ without damaging the other, creating clear dissociation between the two.

The duplex theory, which is a new way of approaching these two streams, classifies ventral as perceptual representations, and dorsal as mediating the visual control of action. The ventral stream acts as the vertical layers of subsumption architecture, while the dorsal stream provides direct sense-action links.

25
Q

How is Piaget’s theory of cognitive development in some ways more embodied than classical theories?

A

Children in their early teens, when acquiring formal operations, use their environment to manipulate objects in the same manner that representational or symbolic thought represents operations. There’s this stage where direct interactions with objects in the world are used to develop formal operations, and are then internalized as symbols.

26
Q

How does Vygotsky utilize the role of social systems in cognitive development theories?

A

The role of social systems is a type of external world, and Vygotsky’s ‘Zone of Proximal Development’ utilizes social support and aid to explain how children develop skills such as problem solving. The development of think is not from the individual to the social but from the social to the individual.

27
Q

What is cognitive scaffolding?

A

The use of the world to support or extend thinking.

28
Q

What is Scribner’s work on mind in action?

A

An examination of practical cognition, Scribner viewed cognitive processes as being embedded with human action in the world - in this regards she studied cognitive scaffolding (“practical problem solving is an open system that includes components lying outside the formal problem - objects and information in the environment and goals and interests of the problem solver”).

Novice versus expert dairy workers: novice workers more often utilized a purely mental arithmetic approach to counting items vs. cases, whereas expert workers took advantage of the environment, relying on visual information (looking at the cases), as well as were more flexible swapping between distribution of scaffolded and mental arithmetic, and did so in a systematic way (when more mental arithmetic was employed it was done to decrease the amount of physical exertion required to complete the order)

29
Q

What is Scribner’s “law of mental effort”?

A

Intelligent agents may be flexible in the manner in which they allocate resources between sense-act and sense-think-act processing; both types of processes may be in play simultaneously, but they may be applied in different amounts when the same problem is encountered at different times and under different task demands.

30
Q

What is a bricoleur?

A

An “odd job man” in French; describes the flexible thinking described by Scribner, where agents alternate between sense-act and sense-think-act in a systematic way.

In Scribner’s observations of the dairy farm workers she observed the bricoleur using (1) external memory and (2) manipulation of data as types of cognitive scaffolding.

31
Q

What is a criticism of the extended mind hypothesis?

A

The line which defines the “mark of the cognitive” becomes ambivalent and that this hypothesis makes no attempts to define the principled differences between cognitive and non-cognitive processes.

32
Q

What is public cognition?

A

A group of agents are embedded in a shared environment, and therefore more than one cognitive agent can manipulate the world that is being used to support the information processing of other group members (ie. a group of individuals operating a ship’s navigation; it is difficult to assemble a procedure that cam be implemented by a navigation team based on the heuristics used by a solo navigator).

33
Q

What is collective intelligence?

A

aka swarm intelligence or cooperative computing; using the idea of the superorganism, colonies can produce complex behaviour that one can predict from knowing the capabilities of individual colony members - a swarm’s components are only involved in local interactions with each other, and result in many advantages. For example, a computing swarm is scalable and therefore flexible (agents can be added or removed from the swarm without reorganizing the entire system). This makes the system robust. Additionally, nonlinearity plays a role in swarm intelligence. In order for a swarm to be considered intelligent, the whole must be greater than the sum of its parts. With a linear increase in agents there can be an exponential observed increased in outputs.

Therefore, collective intelligence displays flexibility and nonlinearity, which are important concepts from connectionist cognitive science.

34
Q

How do you solve the coordination problem (coordinating the varying actions of complex members in a robot collective)?

A
  1. intentional cooperation; uses direct communication between agents to avoid unnecessary duplication
    » the issue with this is that it’s costly to have communication, and that implementing communication makes the actions of the agents specialized
  2. the embodied approach, utilizing stigmergy; robots use a variety of sensors to detect and avoid other robots, locate the box, and locate the goal location; subsumption architecture was used with a fairly simple set of sense-act reflexes; because when one robot changed the environment it affected the actions of all of the other robots,, and because of this (and the built in avoidance behaviour for other robots) the robots were able to do a task together that would not have been possible apart (because the box was too heavy to move with just one robot)
35
Q

What role does history play in the parable of the ant (or at least, the proposed revised version)?

A

Rather than watching an individual ant on the beach, we should arrive at a beach after a storm and watch generations of ants at work; as the ant colony matures, the ants appear smarter, because generations of ants have left their marks on the beach. The ants interaction with residua of the history of its ancestors make it appear smart.

36
Q

What is social scaffolding?

A

The diversity of problem solutions can be generated by utilizing the cognition of others, similar to the historical aspect in the parable of the ants (revised) and the dairy workers communicating and working together.

Scribner says “the need for a greater understanding of the ways in which the institutional setting norms and values of the work group and, more broadly, cultural understanding of labor contribute to the reorganization of work tasks in a given community”

37
Q

Who is Giambattista Vico?

A

The eighteenth century philosopher that links to embodied cognition (another reaction to Descartes); his verum-factum principle is based upon embodied mentality.

38
Q

What is the scientific importance of Grey’s tortoise?

A

His devices were seriously intended as working models for understanding biology; a ‘mirror for the brain’ that could bot generally enrich our understanding of principles of behaviour (such as the complex outcomes of combining simple tropisms) and be used to test specific hypotheses (such as Hebbian learning).

Grey’s tortoises:
» produced complex behaviour with simple devices
» could explore, approach, and avoid
» could respond to hitting an obstacle (which displayed sumsumption architecture)
» could choose
» “mirror dance”

39
Q

What are new wave robotics?

A

They use sense-act cycles to replace sense-think-act ones; they can properly demonstrate the reorientation task with rotational errors.
» one such robot, an evolutionary robot (you 1. define a fitness function, 2. start with an initial population of different control systems 3. use the fitness function to assess each of these control systems, so that those which produce higher fitness values “survive” 4. survivors create the next generation through “mutation” 5. each generation improves until the fitness stabilizes, and the robot should be able to perform the initial function)
» the control system of the robot was merely proximity detectors and motors, not an encoding of the arena shape
» the gentle curving to the left ensures that the robot never hits the short wall
» hits the corner, or the opposite one (rotational error)

40
Q

Describe SLAM and anti-SLAM

A

SLAM = simultaneous localization and mapping; SLAM attempts to answer how agents build a new map of a novel environment and at the same time use this map to determine the agent’s current location. SLAM does so by using representational assumptions to build an internal map.

anti-SLAM is a direct counter-theory; drive < escape < wall-following < feature subsumption structure allows for the robot to display exploratory behaviour; in a reorientation task the robot displays both geometric and featural cue-following abilities, just like the rats do!

41
Q

What are some early experiments into the social environment?

A

They show behavioural loops, feedback behaviour between each other’s actions.

  1. Grey Walter’s tortoises (put both of them together)
  2. Boids: robotic birds which operate on three main principles of (a) avoid collision with nearby flock mates, (b) match the velocity of nearby flock mates, and (c) stay close to nearby flock mates.
42
Q

What are some key concepts of mirror neurons?

A

When the neurons fire, they do so for the entire duration of the observed action, not just at its onset. They are grasp specific. They can be broadly tunes (triggered when a variety of actions are observed) or specifically tuned. All seem to be tuned to object-oriented action (fail to respond when object is absent).

Mirror neurons are believed to be the location to the Theory of Mind concept in the brain (believed, not proven).

43
Q

What is social cognitive neuroscience?

A

An interdisciplinary research program which aims to work out connections between fundamental neurophysiological mechanisms and highly complex social behaviour (and then decide whether the mechanisms are specific to social processes)

44
Q

What are the purposes of social robots (besides being hella cool and sometimes hella creepy)?

A
  1. Provide a medium for studying human social cognition via forward engineering.
  2. To design robots to work cooperatively with humans by taking advantage of a shared social environment.
  3. To explore cognitive scaffolding (aka leverage) - these social robots need a lot of human help to function
45
Q

What is the robotic movement?

A

When a social robot is uncritically accepted as a creature

46
Q

What is mind reading in the context of theory of mind?

A

Having a mental state and representing another individual as having such a state (the latter is a description of mind reading).

47
Q

What are the three (general) competing theories about how humans perform mind reading?

A
  1. Rationality Theory: the intentional stance; mind reading is accomplished via. the ascription of contents to the putative mental state of others; in addition, we assume that other agents are rational; as a result, future behaviours are predicted by inferring what future behaviours follow rationally from the ascribed contents.
  2. Theory-Theory: the position that our understanding f the world, including our understanding of other people in it, is guided by naive theories.
  3. Simulation Theory: the view that people mind read by replicating or emulating the states of others (putting yourself in ones shoes)
48
Q

Describe the simulation theory.

A

Heal: how we predict the behaviour of others is related to how we predict our own behaviour, Predicting our own behaviour is simple because that’s the precursor (the declaration of immediate intention). When predicting other’s behaviour, we predict our own future behaviour as if simulating practical reasoning from the other’s POV (engage in ‘pretend-play’, a key element of which is to take behaviour off-line)

Gordon: exploits the body more, in that it uses it as a replacement for representation; in Gordon’s theory one’s own behavioural control system is employed as a manipulable model of other such systems…because one human behaviour control system is being used to model others, general information about such systems is unnecessary. Rather than having mental representations to play out this ‘in-your-shoes’ scenario, motor representations (such as mirror neurons) are the mental shoes. This theory is more in line with embodied cognition.

49
Q

Describe how embodied cognition is similar and different at the levels of analysis as classical cognition.

A
  1. replaces reverse engineering with a inverse, bottom-up methodology known as forward engineering
  2. builds an interesting primitives into a working system, which is then placed in an interesting environment in order to see what it can and cannot do (ie. embodied cognition starts with implementational and architectural) and creates theories based on how the agent works and how it’s environment contributes (Occum’s Razor)
  3. embodied cognition rejects methodological solipsism and recognizes the crucial contribution of the environment
  4. replaces sense-think-act with sense-act
  5. has all of the same levels of analysis but regards them with differing levels of importance and different types of interpretations
50
Q

What is the uncanny valley?

A

When a robot becomes TOO similar to humans that it violates some type of notion of “biological human” and is no longer as acceptable.