This is the first in a series of blog-posts exploring depictions of androids, robots and AI in popular culture. In these blog-posts I will discuss the philosophical difficulties with these depictions. In this blog-post I will discuss the fictional android Data and his ability to experience emotions and the implications of these experiences for how we interpret the nature of his consciousness.

In an episode of ‘Star Trek The Next Generation’, called the Descent part 1 (Season 6 Episode 26) the android Data experienced the emotion of anger. Data subsequently discussed his emotional experience with his friend Geordi La Forge. La Forge is originally sceptical of Data’s claim to have experienced emotions and asks him how he could know what an angry emotion is. Data asks Geordi to describe his own experiences so he can use them to see if they are similar to his. Geordi’s attempt to describe his emotional experiences turns out to be embarrassingly inept. The dialogue between Data and Geordi is worth expounding in full detail as it reveals a lot about difficulties in describing the nature of emotions:

“Data: I believe I have experienced my first emotion

Geordi: No offence Data, but how would you know a flash of anger from some kind of power surge?

Data: You are correct that I have no frame of reference to use to confirm my hypothesis. In fact I am unable to provide a verbal description of the experience. Perhaps you can describe how it feels to be angry? I could then use that as a reference.

Geordi: Ok…When I feel angry first I feel…hostile!

Data: Could you describe feeling hostile?

Geordi: Well yeah…it’s like feeling belligerent…combative.

Data: Could you describe feeling angry without referring to other feelings?

Geordi: Hmm…no I guess I can’t. I just… feel angry…

Data: That was my experience as well…I simply…felt angry…” (Star Trek: The Next Generation, ‘The Decent’ part 1, 33mins-35 mins.)

There is an element in the above dialogue that puts one in mind of the Socratic dialogues. In the Socratic dialogues Socrates comes across some dupe who claims to understand some abstract concept like ‘Justice’ or ‘Equality’ etc. After a few minutes of being questioned by Socrates we realise that our dupe doesn’t in fact understand these concepts because he cannot even answer simple questions about them. But in the case of the discussion between Data and Geordi we are led to a different conclusion. The brief dialogue is meant to give the impression that although Geordi cannot define the nature of emotions (without appealing to other; equally undefined, emotional terms), he understands the emotional terms based on his immediate experience.

Since it is Data who had just had his first emotion in the above scene it is important to try and understand a bit about him before we can interpret his experience of having an emotional experience. Throughout the series prior to experiencing his first emotion Data is presented an intelligent thoughtful agent who is respected by his colleagues and who is capable of interpreting the behaviour of his colleagues in a largely accurate manner and to use language that is both coherent and (largely) appropriate to the situation.

Given Data’s linguistic proficiency in engaging in communication with his peers, and his ability to interact with his environment, he is typically treated as a conscious member of his tribe. But the question of whether he is a conscious agent is never really dealt with in sufficient detail. Though there is an episode where he has dreams and this is indicative of his being conscious; I will discuss this in more detail later on.

First I want to briefly discuss an early episode of Star Trek The Next Generation where Data is put on trial to defend his status as a person as opposed to mere property. In the trial it is argued that a creature must meet three criteria in order to be considered a sentient agent. They must be intelligent, self aware and conscious. The first two criterion are met when Data indicates intelligence through his performance in various tasks, he is judged to be self aware because he can verbally describe the scenario he finds himself in, on the question of consciousness it isn’t proven he is conscious it is merely remarked that it is as hard to demonstrate that other humans are conscious as it is to demonstrate Data is.

If we take Data at his word that he is has experienced the emotion of anger then we will be forced to admit that he is conscious. However there were moments earlier in the series that would cause one to doubt that diagnosis. Firstly Data doesn’t appear to feel any pain, thus he has at times had his head removed, his arm removed, and doesn’t indicate any pain or discomfort whatsoever. Along with not feeling pain, he doesn’t seem to experience pleasure; thus while he has had sex he doesn’t associate it with any pleasant sensations. Though it should be noted that while there is no evidence he experienced any sensations when having sex; he did afterwords describe the experience as a meaningful one for him.

The question is can a creature who is incapable of experiencing pain or pleasure have conscious experiences? If no aspect of our environment gave us pleasure or pain; in what sense would it make sense to argue that a creature was conscious? In one sense one could argue that consciousness is simply awareness of something. For a person to be aware of x it doesn’t intuitively seem necessary for one to experience pain or pleasure. It would appear to be possible to view a red object as a red object without this experience involving experiences of pleasure or pain.

To intuition relies on the representational theory of the mind; which was made famous from Descartes argument from error. Descartes argument relies on demonstrating that the world as it is revealed by science isn’t the world of experience. Thus, for example, a stick in water can appear to be bent, but in reality the stick is actually straight. With a disjunction between how the world appears to us, and how it really is; the stage was set for a representational theory of the mind. Descartes theorised that the world as it appears to our mind is a representation which we use to make sense of the world but it is not a direct experience of the real world. Not many people followed Descartes in his dualistic way of conceiving of this issue but the majority of scientists and philosophers follow him in accepting that the argument from illusion leads to a representational theory of the mind.

Again consider Data and his experience of the world. A person walks by him wearing a red top; light reflects off the red top and hits Data’s eyes the light is registered by the eye and information is transmitted from his eyes to his positronic brain, and his brain (somehow) creates a representation of the red object. It is this representation that Data experiences. So if we accept this story[1]; we can argue that Data had conscious experiences in the form of representations. There is some support for this interpretation in the Star Trek episode:

https://youtu.be/fWE2cHiEz6E

In the above scene we are presented with Data’s dream from a first person perspective. Now a dream is an example of Descartes argument from error. In a dream we purportedly experience a world around us; however since we are really asleep in our beds, the world we see isn’t real but is rather a representation of reality that is fooling us. Now given that in the above episode Data is portrayed as being capable of dreaming it seems inescapable but to conclude that he has conscious experiences[2]. But it is a strange disembodied kind of consciousness.

Data is portrayed as having conscious representations of the world which contain rich qualitative experiences of colour, sound, shape etc. But other aspects of his behaviour seem unconscious or reflexive. As we saw above Data seems to have absolutely no pain receptors. His body can obviously register aspects of his environment and respond appropriately to it. But as portrayed in the show he doesn’t seem to experience any conscious proprioception as he moves about his world. In a sense Data’s body is similar to the body of the robots currently being built by Boston Dynamics:

https://youtu.be/hSjKoEva5bg

The above robot, like Data, is adroit at moving around its environment and like Data his skilled movement is unaccompanied by any subjective experience. Of course to a degree we are all like Data or a Boston Dynamics robot as we move around our environments. People who have a severe stroke who are trying to re learn how to walk quickly discover how much of our movements around our world rely on non-conscious mechanics that have to be re-learned post stroke. But there is a difference and it is one nicely captured by Heidegger’s distinction between the ‘ready-to-hand’ and the ‘present-at-hand’. Heidegger notes that when engaging in our everyday activities our movements are pre-thematic, they are in the background and we don’t notice them. When working in this state we find things in our environment ‘ready-to-hand’. However, when things go wrong with our relation to our environment we become aware of ourselves and the objects we are interacting with in a more explicit and theoretical way called ‘present-at-hand’.

In a lot of cases our movement from ‘ready-to-hand’ to ‘present-at-hand’ involves things like a feeling of pain or a feeling of frustration. So, for example, a guy working in a warehouse who is moving things in a particular way on a daily basis will not be aware of the movements till he starts suffering from back pain; once he starts experiencing this pain his interactions with the objects he is moving will switch from ‘ready-to-hand’ to ‘present-at-hand’. The objects he is interacting with will suddenly acquire a particular salience and he will need to be more conscious in his movements when interacting with the objects.

Likewise, if a worker has a well worn set of behaviours when engaging with some aspect of his environment and these behaviours cease to work this will inspire some emotions. The worker may feel frustrated or curious and will have to step back from his behaviour and form a new way of interacting with the objects in his environment.

The situation with Data though precludes such switches from ‘ready-to-hand’ to ‘present-at-hand’, Data is incapable of experiencing pain, and he is incapable of experiencing frustration. So if Data was a worker in the above situations he would have little reason to switch to conscious deliberation when things go wrong. Data is constantly portrayed as experiencing curiosity and wonder. But the show never offers an explanation of why a creature who is supposedly devoid of emotions is capable of experiencing curiosity.

Another strange aspect of Data as a concept is his development. He is portrayed as an android who was built with an adult body and who was programmed to move around his environment, speak respond etc. All of this behaviour was built in; he just used these built in competencies to learn the nature of the world he lived in through interaction with humans over a number of years. So Data never experienced life in a womb; or as a child entirely dependent of his care givers; nor did he feel the human innate emotional bond with parents and peers. He never felt embarrassed, or angry, or exhilarated when interacting with childhood friends, parents, neighbours etc. He never went through puberty and the emotional changes it brought about.

So with Data we have a weird combination. He is supposedly capable of conscious representations of things such as shape, colour, size etc. But he doesn’t experience pain, while he can touch things and interact with them; such interaction appears to be on the level of a Boston Dynamics robot (with the added ability to theoretically interpret his own movements). Whatever consciousness Data supposedly has it seems to be entirely disembodied and like a magical add on to his mechanical behaviour.

This leads us back full circle to Data’s conversation with Geordi. Geordi cannot explain his emotional experiences in a satisfactory manner. But he has a lifetime of experience of learning emotional words and keying them with bodily movements and various experiences. He has a life time of shaping his emotion words to his embodied experience and using them in a way that his similarly constituted peers can understand. Data on the other hand has always been able to use emotion words, but because he had no emotions there was no fit between his words and his experiences. Given Data’s weirdly disembodied Cartesian Nature it is unlikely that even if he suddenly did have new experiences (which he would categorise as emotions), they would line up with human emotions. Data if he were suddenly given emotions would be like a person who was blind for his entire life suddenly being given his sight. Even with the ability to see the ability to interpret distance, size etc would take years of learning to be perfected. And there is a strong possibility that it would never be entirely perfected.

[1] And there are plenty of difficulties with the story; firstly while we have good story of the neural correlates we have no idea how those correlates produce our experiences in the way they do; secondly any representational theory of mind runs the risk of an infinite regress.

[2] Dennett’s 1976 paper ‘Are Dreams Experiences’ argues that dreams are not something we directly experience but rather stories we rationally reconstruct after waking. But in the fictional world of Star Trek we are shown Data directly experiencing his dreams; so within this fictional world Dennett’s concerns are mute. Though a philosopher living in the fictional trek world could argue a la Dennett when Data reported his having dreams.

    Introduction

In his recent excellent book ‘This View of Life’ David Sloan Wilson offers a compelling case for the utility of evolutionary thinking in all areas of life; in particular he makes a masterful case for how evolutionary thinking can improve our over-all standard of living and make our world a better place. In future blog-posts I will discuss Sloan-Wilson’s recommendations for using evolutionary thinking for public policy. In this blog-post my focus will be narrower. I will examine Sloan-Wilson’s very brief discussion of B.F. Skinner. Like every evolutionary psychologist who criticises Skinner, Sloan-Wilson uses the trope of Skinner being a blank slate theorist.

Sloan-Wilson on Skinner and Behavioural Science

A lot of evolutionary thinkers who criticize Skinner (and behavioural science in general), don’t demonstrate any understanding of the subject; but Sloan-Wilson is different. His work with behaviourists such as Tony Biglan and Steven Hayes has given him an understanding of the subject far beyond the likes of Pinker or Fodor. Sloan-Wilson acknowledges the important work in applied science that has been done by behaviourists; e.g. work by ABA practitioners with Autistic children, or the good behaviour game which he tries to link with his own evolutionary understanding of science. Nonetheless Sloan-Wilson is critical of behavioural science in general (in particular he is critical of Skinner’s radical behaviourism).

Sloan-Wilson is complementary towards Watson and Skinner’s criticism of the introspective psychology that preceded the behavioural revolution. However, he is critical of Skinner for only focusing on half of Timbergen’s famous four questions. Timbergen noted that when trying to understand the behaviour of an organism we should do so by asking four different types of question (1) history, (2) function, (3)  mechanism, (3) development.  On Sloan-Wilson’s picture Skinner ignored question (3) and (4). As a result of this Sloan-Wilson uses the tired old label of blank slate theorist to categorise Skinner’s position.

Sloan-Wilson’s criticisms of Skinner have some validity. Skinner did argue against explaining human behaviour in terms of inner mechanisms or cognitive models. Firstly it is important to note that being a sceptic about the value of cognitive maps wasn’t a behaviourist dogma. Behaviourists, such as Skinner’s contemporary, Edward Tolman used cognitive maps in their explanations of human behaviour. Nonetheless, Skinner is the name contemporary thinker’s associate with behavioural science; so Wilson is correct to point out Skinner’s scepticism re- the importance of cognitive maps. Furthermore, throughout his entire career Skinner showed little interest in developmental issues.

The question we have to ask here is whether Skinner erroneously ignoring questions 3 and 4 make him a blank slate theorist? In order to evaluate this question it is worth considering a comparison that Fodor and Piattelli-Palmarini[1]made in their infamous book ‘What Darwin Got Wrong’.

 

 

Fodor on the analogy between Skinner and Darwin

F and P argued that the logic used by behaviourists re- operant conditioning and the logic used by Evolutionary theorists was identical. They noted that there was an inconsistency in the way that Skinner and Darwin were treated; people argued that the Skinner’s logic lead to an easy refutation; while Darwin despite using the same logic didn’t receive the same treatment.

The first point of comparison they make is in terms of what they call population thinking. They claim that a way to think of the theory of natural selection[2] is as a theory of how phenotypic properties of populations change overtime in response to ecological variables (‘What Darwin Got Wrong p.3). They define Operant Theory[3] in a similar way ‘OT is also plausibly viewed as a black box that maps a distribution of traits in a population at a time (a creatures behavioural repertoire at that time), together with a specification of relevant environmental variables, the creature’s history of reinforcement; onto a succeeding distribution of traits.

They argue that the TNS and OT have 6 basic untenable feature in common: (1) Iterativity: ET provides no bounds on the type of phenotype possible OT provides no bounds on the variety of behavioural profiles which can be created through conditioning.(2) Environmentalism: ET and OT abstract from endogenous variables, claiming that the phenomenon of evolution on the one hand and psychology on the other are largely the effects of environmental causes (3) Gradualism: ET argues that new phenotypes emerge gradually, OT argues that learning is a gradual process of stimulus response conditioning (4) Monotonicity: ET and OT are one factor theories. For ET selection does all the work. For OT conditioning does all the work. (5) Locality: Both ET and OT are local processes and are insensitive to mere hypothetical contingencies (6) Mindlessness: ET doesn’t postulate God to do the work and OT doesn’t postulate Mind to do the work.

Fodor and Piattelli argue that the evidence they provide in their book shows that evolutionary theory as defined according to the above six principles cannot do what it purports to do. And in this sense they argue both OT and ET are hopeless theories the only difference is that people understand that OT is hopeless; yet inconsistently they remain wedded to ET.

Internal constraints: The authors begin this section with a claim that standard Neo-Darwianists are environmentalists by definition. By this they mean that standard Darwianian theory thinks that changes to a phenotype are largely driven by environmental contingencies. Their primary aim in this section is to show that contemporary wet biology is telling a story of innate constraints which are at odds with the neo-Darwinian story. It is worth noting that the authors they cite in this section do not agree with the use F and P make of their work. Furthermore, most neo-Darwinians would deny that they are environmentalists in Fodor and Piattelli’s sense. So they would argue that Fodor and Piattelli are attacking a straw-man.

Bearing all of this in mind lets now review the evidence they cite. The first thing they cite is the concept of Unidimentionality. Unidimentionality is supposedly standard story in the neo-Darwinnian theory. On this picture NS plays the primary role in the theory of evolution, the role of internal sources of variance, and internal constraints is said to play only a marginal role. To prove this point they cite Earnest Mayr’s book ‘Animal Species and Evolution’ as an example of such ultra-selectionist attitudes. F and P claim that discussion of the evolution of the eye nicely illustrates the neo-Darwinian emphasis on NS as the primary source of design in species. It was claimed by most neo-Darwinian theorists that the evolution of the eye emerged several times independently and convergently across species. In his Darwins Dangerous Idea Dan Dennett referred to the evolution of the eye as a nice trick, something that was bound to be selected in any form it occurred in. Dawkins has made similar claims. F and P pointed out that the discovery of master genes for eye development (Pax 3, Pax 2, Pax 6, and Dach) across vastly different classes and species has shown the neo-Darwinain view to be incorrect. The next topic they consider is beanbag genetics. Here they basically argue that selection for a particular gene rarely, if ever occurs, and this is because of the convoluted packing of genes in chromosomes. Their critique of beanbag genetics is a pointless because nobody believes it anyway.

One of the key factors they believe counts against the neo Darwinian view is the existence of Internal Constraints and Filters. The discussion of internal constraints and filters involves an appeal to results in the evo-devo revolution. Again it is worth noting that most people working in evo-devo consider themselves a part of neo-darwinanism and would not accept the conclusions drawn from their work by Fodor and Piattelli. According to Fodor and Piattelli, the standard neo-Darwinian picture abstracts away from the all effects of development on visible traits (p.27). They stress that the evo-devo revolution shows that this development not only cannot be abstracted away from, it is key to the process of evolution. The argue that it has been shown in the lab (1) phenotypic convergence is, more often than not the result of developmental constraints, (2) Also they cite the fact that experimental evidence (Ronshaugen 2001), has shown that terminal forms can differ in massive ways as a result of slight variations in the regulation of the same gene complexes/or the timing activation of such complexes (pg 30). This shows that contrary to neo-Darwinian claims evolution is not primarily driven by exogenous factors but by internal developmental constraints. They spend the rest of the chapter outlining a series of facts which they claim further develop their point. Throughout section 1 they are merely attacking a straw-man, because most evolutionary theorists do not deny what they are claiming. Though it is true that a lot of pop science is guilty of making claims of the type they critique.

In chapter two and three F and P argued that there are internal constraints which limit the importance of selection, and they considered how if at all selection could operate given these limits. They claimed that in response to the evidence reviewed in chapters two and three neo-Darwinists have expanded its scope and invoked other kinds of natural selection. This chapter is an attempt to provide more problems for neo-Darwinism. The first problem they consider is the phenomenon of adaptation without selection, Fodor and Piattelli summarise the point as follows:

“The point to keep your eye on is this: it is possible to imagine serious of alternatives to the traditional Darwinian consensus that evolution is primarily a gradualistic process in which small phenotypic changes generated at random are then filtered by environmental constraints. This view is seriously defective if, as we suppose, the putative random variations are in fact highly constrained by the internal structures of evolving organisms. Perhaps it goes without saying that if this internalist story is true, then less work is left for appeals to natural selection to do.” (What Darwin got Wrong p. 54).

They provide eight pieces of evidence which they think support their conclusion:

(1)   Gene Regulatory Networks: Building from the work of E. H. Davidson (2006), they argue that gene regulatory networks are at work in the development of the organism. These gene regulatory networks are modular in nature (in other words they form compact units of interaction which are separate from other similar units). The important point about these regulatory networks is that they are supposedly responsible for the development of the bodily structures of animals. This happens because large effect mutations acting on conserved core pathways of development. They claim that this process makes it virtually impossible to argue that particular isolated traits are selected for.

(2)   Entrenchment: They claim that this acts as an engine of development and evolutionary change, and as a constraint (ibid p.43). Some evolutionary factors may be highly conserved and protected against change. They offer very little evidence of their views at this point merely a promisary note to develop the point in the next chapter.

(3)   Robustness: This is the persistence of a trait of an organism despite developmental noise, environmental change or genetic change. This robustness is important for the stability of phenotypic change despite genetic and non-genetic variation. They cite the work of Wagner (2008) which claims that it is only the additive component of genetic variation which responds to selection.  Fodor and Piattelli argue this fact should make people wary of accepting the neo-Darwinian view that selection is the primary vehicle of phenotypic variation.

(4)   Master genes are our ‘Masters’: They make the now well established point that many genes are indissociably controlled the same ‘master gene’. Therefore if a mutation affects a master gene (and is viable) it will affect all of the genes the master gene controls as well. They link this to Gould’s famous paper on spandrels. They briefly discuss how the evolution of language may not be explicable by a simple adaptative story in terms of selection for communication. Using facts about master genes they argue that language may have been a free-rider, which was selected because some mutation in the master gene Otx. They claim that this story is not even considered because of allegiance to an ultra adaptationist model. I do not agree with this claim there has been ample debates on this topic. See Hauser, Fitch and Chomsy 2005 and reply by Jackendoff and Pinker 2006. However evaluating this debate would take a long discussion of linguistics which is beyond the scope of this discussion.

They go on to further discuss things like developmental modules, coordination, morphogenetic explosions, plasticity and the (non-transitivity) of fitness. All of these facts are well known in the literature and it is unclear to me at least why they believe these facts pose a major problem for evolutionary theory. They do pose a problem for the caricature of evolutionary theory they present at the beginning of their book but not for evolutionary theory as it is actually practiced.

They also consider ‘Laws of Form’ as an argument against the standard Neo-Darwinian Story. They discuss the work of thinkers like Stewart Kauffman, Stuart Newman, and Lewis Wolpert who have all discussed the important topic of laws of form and self-organisation. Fodor emphasises how this research shows that we need to discover what forms are possible for an organism to take before we attack the question of how selection can act on these possible forms. These constraints on possible forms are shown in things like non-genomic Nativism discussed by people like Cherniak. Cherniak details computational constraints on brain anatomy which he claims are derived from physics for free; hence we do not need natural selection to explain some of the structure of the brain. F and P also discuss the work of James Marden who has detailed physical constraints on possible animal locomotion. Their discussion of laws of form is extremely interesting but again it is hard to see that it really poses any problem for the standard neo-Darwinian picture. There really is nobody, and I mean nobody, who denies that there are physical constraints at work in evolutionary theory. They are correct to note that pop evolutionary writers sometimes ignore these physical constraints and focus entirely on selection. So, if F and P were merely warning against this type of mistake, then their point would be well made, but it should be obvious that their arguments do not have any bearing on neo-Darwinian theory when construed correctly.

The Relevance of Fodor’s analogy for Sloan’s treatment of Skinner

F and P’s argument against Darwinism was largely ridiculed as ignorant nonsense by evolutionary scientists. The correctly noted that their book[4] was an accurate description of work that was ongoing in the discipline; but that it was pure fiction that this work refuted Neo-Darwinian theory. On the contrary, the work was a part of the modern Darwinian Synthesis.

The relevance of this digression to the present discussion F and P’s treatment of OT and ET was almost identical. F and P treated ET as an environmentalist theory that focused entirely on selection and ignored constraints that limit the power of selection. Likewise F and P treated OT as a theory who overrated the power of operant conditioning and ignored the limits innate factors would place on such conditioning. F and P were right about one thing though there is an inconsistency between the way OT and ET were treated. F and P were laughed out of court for their caricature of ET. But most theorists seem to accept their caricature of OT as factual. Even a thinker as sophisticated as Sloan Wilson seems to accept the caricature:

“Eventually, the limitations of behaviourism became apparent. Organisms were not entirely blank slates in their learning abilities, and techniques were becoming available for studying how the mind works…” (Sloan Wilson ‘This View of Life’ p. 100)

Skinner of course never claimed that organisms were blank-slates. Throughout his career Skinner argued that two primary sources influenced shape human behaviour (1) The contingencies which shaped the species (resulting in genetic constraints), (2) The contingencies of reinforcement that shape the organism throughout it’s life. (See About Behaviourism p. 168, Beyond Freedom and Dignity p. 14, Verbal Behaviour p. 162, Science and Human Behaviour p. 26 etc.)

Below are some quotes from Skinner where he explicitly denies being a blank slate theorist:

“As a result it is part of the genetic endowment called “human nature” to be reinforced in particular ways by particular things.” (Beyond Freedom and Dignity p. 104)

“Just as we point out the contingencies of survival to explain an unconditioned reflex, so we point out to ‘contingencies of reinforcement to explain a conditioned reflex” ( Skinner About Behaviourism p. 43)

“The task of a scientific analysis is to explain how the behaviour of the person is a physical system is related to the conditions under which the human species evolved, and the conditions under which the individual lives” (Skinner ‘About Behaviourism’ p. 14)

The above quotes and references indicate Skinner was far from a theorist who believed in a blank-slate. In fact he emphasised that the conditions under which a species evolved would heavily influence what they would find reinforcing. It is true that he was more interested in environmental factors than the innate constraints but it is unfair to characterise his work as blank-slate. Attempts to reduce Skinner to a blank-slate theorist are as unfounded as attempts to claim evolutionary scientists think natural selection is the only factor that plays any role in evolution.

[1] Hence forth Fodor and Piattelli-Palmarini will be referred to as F and P.

[2] Henceforth the Theory of Natural Selection will be referred to as TNS

[3] Henceforth Operant Theory will be referred to as OT.

[4] I am here only discussing the first part of F and P’s book. The second half was a conceptual argument which while interesting has little bearing on the present discussion so can be ignored here.