Inaugural Lecture Series: Professor Jose Prados video transcript

The title screen appears. A blue background is shown with the University of Derby three hills logo in white in the middle of the screen. The University of Derby's name in white is directly below. White text on the screen reads below:

Professorial Inaugural Lecture Series:

The Evolution of Learning: Towards a Phylogenetic Epistemology by Jose Prados

The title screen fades out and the presentation title slide is showing. Covering the slide is an angled image of the University of Derby’s Kedleston Road Campus and the sky. In the top right corner, a white title reads ‘Jose Prados Inaugural Lecture’. The University of Derby 3 hills logo in white is placed in the top left corner. In the bottom right corner is a live video of the inaugural lecture taking place in a lecture hall at the University of Derby. This will be shown throughout the entire presentation.

[Paul] Good evening, ladies and gentlemen. For those who don't know me, I'm Professor Paul Lynch and as chair of the University Professorial Council I would like to welcome you to the latest series of inaugural lectures here at the University of Derby. This time we've got Jose Prados, who is the head of psychology at the School of Psychology at the University of Derby.

I would like to introduce Dr Denise Baker, who is the PBC dean interim for the College of Health, Psychology and Social Care to introduce Jose.

[Denise] Thank you, Paul. Welcome everybody, It's my pleasure to introduce Jose to the audience. Jose was born in 1967, the third of four Prados children in Barcelona in Spain. He had an interesting childhood that involved the other presence of his maternal grandmother and 60 cousins.

Being asthmatic, he benefited from being away from the big city, so he spent some years in a Catholic boarding school in the Pyrenees near Andorra. Curiously enough, his parents were members of the then illegal Communist Party, so diversity and contradictions were at the core of his education. In the mid 1980s, he started his degree in psychology at the University of Barcelona. On the first day of the autumn term, as a fresher one of his lecturers (Dr Victoria Chamizo) introduced her module and told students that they would be assessed on the contents of a particular book, ‘Animal Learning and Motivation by Roger Tarpey’. She also asked for volunteers to work in the Animal Psychology Lab and Jose instantly stepped forward, and this started his lifelong collaboration with members of the team.

During the first year, he wrote learned, (and that's not a recommendation, it's just a statement). he wrote Learned the contents of the Tarpey book and scored a ridiculously high 98.5% in animal learning. A true oddity, Somehow, he wasn't as brilliant in other areas.

That was the beginning of a long and colourful disciplineship to Dr Victoria Chamizo, who eventually became Jose’s doctoral supervisor. Chamizo (one of the leading experts in spatial learning) had a very close collaboration with Professor Nick McIntosh, the head of the Department of Experimental Psychology at Cambridge.

While still an undergraduate, Jose developed with Professor Chamizo and Professor Macintosh, the Morris Pool facility in Barcelona. He later used this facility for his PhD project on pre-exposure effect on a navigation task with rats. His interests were still diverse and he was simultaneously training as a family therapist. During his PhD, Jose benefited from long annual studies of several months in different UK universities where he was lucky enough to collaborate with John Pierce in Cardiff and Jeffrey Hall in York, working on the associative analysis of spatial learning and perceptual learning.

In 2000, Jose was amongst the first students in Barcelona (and possibly in Spain) submitting a doctoral thesis as a compilation of published papers. Following the award of his PhD, Jose held fixed term lectureships in his home department. In this period, he established himself as an independent researcher funded by the Spanish Ministry of Science but was growing increasingly dissatisfied with the lack of prospects in Spain, where even prominent researchers were only offered short term contracts at the time.

In 2004, Jose, his wife Noria and baby Simon decided to make England their home. Jose was offered a lectureship at the University of Leicester, where he spent 17 fantastic years teaching and researching. He held a diverse range of roles including Head of School of Psychology from 2017 to 2021 and the Prados family welcomed two more Leicester born children, Martin and Leon.

in 2021, attracted by the immense talent and reputation of the psychologists here at Derby, He seized the opportunity to become the head of a newly established school of psychology and subsequently a new adventure has begun. Ladies and gentlemen, I am very pleased to introduce to you, Professor Jose Prados.

[Jose] I would like to take the opportunity to respect convention and thank everybody in my life, past, present and future for the impact they are having in my work and in my personal life and I would say that both in private business and in academia they are all achievements. But mine in particular are the product of a collective effort, so I'm really thankful to you all and I'm really thankful for you to come today to this lecture.

The slide changes to a white background. The right side of the slide is covered by an angled image of the University of Derby’s Friar Gate Campus and the sky. Black text on the left side of the slide reads ‘The Evolution of Learning. Towards a Phylogenetic Epistemology’. The University of Derby 3 hills logo in black is placed in the top left corner.

[Jose] My title is ‘The Evolution of Learning: Towards Phylogenetic Epistemology’. I like this title because most people don't have a clue about what it is about and that includes myself. I guess that the subtitle is redundant and actually means the same thing as the title. In a nutshell, I would say that our interest is to explore how learning emerged in evolution, how it evolved, became gradually more and more complex and how is it distributed in that phylogenetic tree? That would be my talk. Now you can go to sleep if you wish and save the next 35-40 minutes *Laughter*

The slide changes to a white background. The University of Derby 3 hills logo in black is placed in the top right corner. On the right side of the slide, circles with different images within are placed. The images contain a Crow, Goldfish, Bee, Gorilla and an image of a historic man. Black text on the slide reads:

Evolution of Cognition and the Emergence of Consciousness
In recent years, we have witnessed a resurgence of the interest for the concept of consciousness.

[Jose] So, this is it, evolution. This is about evolution of cognition and one of the focuses today is going be on the emergence of consciousness. So, we have been interested in consciousness since forever but nowadays it has stopped being just a philosophical issue, it is now the subject of empirical investigation 

More black text appears on the slide that reads:

From an evolutionary perspective, we consider cognition and consciousness as products of biological evolution.

[Jose] From an evolutionary perspective, we consider cognition and consciousness as a product of biological evolution and the main question that we want to answer

More black text appears on the slide that reads:

The main question we want to answer is who (what animal species) is conscious.

[Jose] is who and by who, I mean what animal species actually have a conscience and have the capacity to experience a consciousness state? It may be just the humans, it could be humans and great apes, it could include some birds, some fish. The question is whether animals, relatively simple in their brain architecture are capable of experiencing conscious states.

The definition of consciousness is, of course, of relevance because if you define consciousness in a very constrained manner, you are reducing the possibility of finding evidence to only within humans. If you are too liberal, then you become a pan-consciousness theorist and you end up claiming that inanimate objects and the universe also have their conscience, and this is our biggest threat. I suppose that we will stay a little bit in the middle.

A grey box is added onto the slide with white text inside that reads:

We use Pavlovian tasks to help assess whether animals meet the criteria to be considered conscious.

[Jose] So, the way in which we are going to address the presence of consciousness throughout the animal kingdom and beyond is by the use of Pavlovian tasks. I've been introduced more or less as a learning theorist but the only thing that I know how to do is Pavlovian conditioning and my plan is to use Pavlovian conditioning tasks to assess the cognitive abilities of different animals and see whether they actually meet the criteria to be considered conscious.

The slide changes to a white background. The University of Derby 3 hills logo in black is placed in the top right corner. Black text on the slide reads:

What is consciousness?

Consciousness is the capacity to have subjective experience. We can distinguish different forms of consciousness

[Jose] So, what is consciousness? This is the first question that we have to address. Consciousness is the capacity to have subjective experience and we can distinguish different forms of conscience.

A blue box appears on the slide with black text that reads:

Sentience

The capacity to experience pain, distress and/or harm
A sentient being is “conscious” in the most elemental sense of the word
It need NOT be able to consciously reflect on its feelings, or to understand the feelings of others.
To be sentient in simply to have feelings 

[Jose] Firstly, some beings can experience sentience and this means that they can experience feelings, but they have consciousness only at a very low level. It's a very elemental form of consciousness; You can have feelings, but you might not be able to consciously reflect on your feelings and of course, you might not be able to be aware of the feelings of other people or other individuals. So sentience is a first step I suppose in the evolution of the world's consciousness.

A dark blue box appears with white text that reads:

Consciousness

The capacity to have subjective experience (subjective point of view on the world and its own body 

[Jose] Being conscious is a different thing, it’s more sophisticated, it gives you the capacity to have subjective experience, some agency, you have a sense of your individuality, you can distinguish yourself from the background and from the world and you have found a subjective point of view of yourself and the world. If you are conscious in that way, you might be able to display goal directed action, so to behave with purpose and you might be able to be aware of the feelings, the knowledge and the intentions of other individuals.

This is the theory of mind. So, if you are capable of recognising that others have feelings and knowledge and intentions, then you are likely to behave in a compassionate manner and behave with humanity. So, consciousness is something more complex than mere sentience and defines what being human is about. But we are sharing that (this is my hypothesis) with other beings.

A purple box appears with white text that reads:

There must be a line separating systems with and without conscious experiences/with and without feelings … Where is the line?

[Jose] The question is, consciousness must have evolved, must be the product of biological evolution and we need to identify a line that separates those that can experience consciousness states from those that cannot, or even, discriminate between those that can have feelings and those that cannot have feelings depending on their biological machinery and the cognition that they have evolved. Where is the line?

The slide changes to a light grey background with black text that reads:

Markers for consciousness

Ginsburg and Jablonka have identified Unlimited Associative Learning (as opposed to Limited Associative Learning) as the Evolutionary Transition Marker for Conscience (e.g., Ginsburg & Jablonka, 2019).

To the left of the text is an image of the book cover for ‘The Evolution of the Sensitive Soul’ by Simona Ginsburg and Eva Jablonka. Underneath the text are a series of images of different organisms, starting from small (atoms) to large (humans).

[Jose] To help with establishing where the line is to discriminate who qualifies for a conscience or not, we can use the work of Simona Ginsburg and Eva Jablonka. They've been working on that issue for the last 25 years, perhaps all their lives. They came up in 2018 with this book which is a big one and they make a number of proposals which are really useful to organise the field of invertebrate learning. In this book, what they suggest is that something that they refer to as unlimited associative learning can be an evolutionary transition marker for conscience. That means that if you assess the behaviour of a particular animal species and they are capable of showing limited associative learning, then these animal species would qualify for having conscience and the ability to experience conscious states. So, the line could be here separating humans from the rest of the organisms, and it would be the new hypothesis.

We know that we have conscience, and we can identify different sorts of learning; Very simple habituation and sensitisation, limited associative learning and unlimited associative learning. By assessing the cognitive skills of different animal species, we can draw the line between species. Perhaps the vertebrates and the octopus (which has a special status) are capable of consciousness? Perhaps, it is the animals with a brain and a central nervous system? And perhaps it can go beyond that, but I would be reluctant to go further. So, beyond this line, there would be animals with a non-centralised nervous system (without a brain). *inaudible* This is what they are proposing.

The slide changes. Black text on the slide reads:

Pavlovian conditioning

Below the text, a diagram is shown of the conditioning method using dogs and bell ringing. To the right of this, a black and white image of Ivan Pavlov.

[Jose] So, before we continue, I already said that my plan, is to use Pavlovian conditioning for everything simply because this is what I know how to do as well as a little bit of instrumental conditioning, but today I'm focussing on Pavlovian conditioning.

Remember that Pavlovian conditioning is a form of learning which allows individuals, animals for example, to learn about the social structure of the environment. They can identify signals and relevant outcomes in their environment. So, this is great because then they can change their behaviour during their lifespan and they become more adaptive. They can adapt to the changing conditions of the environment and they have more chances for survival and reproduction. So, Pavlovian conditioning is a very powerful tool to have for an individual, animal or human. Some people even claim that associative learning, Pavlovian conditioning and instrumental conditioning has been the main driver of evolution. That, at some point in the Cambrian period for example, the emergence of associative learning was responsible for the huge diversification of the animal kingdom that we can still see today.

So, the paradigm that Ivan Pavlov came up with was playing with dogs, and basically, with dogs we know that they are very interested in food and that when they are presented with food, they would naturally salivate, they will automatically respond to these biologically relevant stimuli. When they are presented with a bell however, They can react as this is relatively neutral stimulus that would provoke an auditive response, but the animal will not drool in the presence of the bell. So, the Pavlovian paradigm consists of pairing the neutral stimulus with a biologically relevant stimulus (the unconditioned stimulus) and after a few pairings, if you present the initially neutral stimulus (the bell), then the animal will salivate. The notion is that during conditioning, the animal learns an association between the neutral stimulus and the unconditioned stimulus (the food) and at the time of testing when you are presenting the bell the animal activates a mental representation of the food and this is what provokes the conditioned response of drooling. So this is what we are going to use in our work.

The slide changes and black text appears that reads:

Limited vs. Unlimited Associative Learning

To the left, a black and white image of Donald Hebb, with an effective learning diagram above.

[Jose] We now need to distinguish between limited and unlimited associative learning. This is of importance because, according to Ginsburg and Jablonka, Unlimited associative learning would be an evolutionary transition marker for consciousness. And how do we do that? Well, I don't really know. I keep reading their papers and their book and I always struggle with the way in which they define this. So, I'm going to do it in my own way and one way of distinguishing between limited and unlimited associative learning is by reference to contiguity. So, in the Pavlovian paradigm, for example, one of the conditions that can be observed to ensure that there is learning is that the conditioned stimulus and the unconditioned stimulus are presented close together in time, there needs to be some contiguity between them. The closer the hearing time, the faster the animal would learn an association between them. 

More black text appears that reads:

Limited Associative Learning can take place in a simple manner where mere contiguity between events is seen as a sufficient condition for learning.

Some refer to this as Hebbian learning (e.g., Prados et al., 2013).

[Jose] Some people have referred to this kind of learning as Hebbian learning. I don't know whether this is entirely fair, but at the time I was writing a paper in 2012 or so and I distinguish between simple and complex Pavlovian conditioning, which was brilliant, but the reviewers didn't like it so they suggested using Hebbian and based on the notion that when the light and the shock, for example, e presented that some neurones in the brain will be activated and according to the quotation, neurones that fire together wire together. I don't think that Hebbian actually said that, but the idea is that if there is co activation of the neural centre responsible for the processing of the light and the shock, then an association will establish between them automatically.

So, contiguity certainly works, if you present two things close in time, they will become associated. But this is not always true and this is the second time that we are addressing contiguity today.

More black text appears on the slide that reads:

Contiguity is often not a sufficient condition for learning: the involvement of selective attention or integration of information would define Unlimited Associative Learning.

[Jose] Contiguity is often not a sufficient condition for learning. In many cases, the learning in a Pavlovian task is modulated by other cognitive devices. For example, selective attention can be involved in filtering the information, or the animals can integrate information from different learning episodes, which challenge the principle of contiguity. So, let me show a couple of examples just to demonstrate that contiguity is not always necessary to obtain good Pavlovian conditioning.

The slide changes. Black text on the slide reads:

CS-US contiguity is not a sufficient condition for learning - I

Latent Inhibition (Lubow, 1965): Familiarity with the CS retards conditioning

Below this text is a repeated image showing effective learning, using ‘light’ in a blue circle with an arrow drawn to a ‘lightning bolt’ in a red circle.

[Jose] So, the first one is latent inhibition. It was discovered by Robert Lubow back in the 50s, and this is based on an experiment published in 1965 that was done with Goat. In that case, two groups of animals were trained in a standard Pavlovian conditioning task. A light was the signal for a shock (unconditioned stimulus) which was delivered to the rear leg of the animal, and this shock resulted in the flexion of the leg. So, with sufficient training, the light would acquire the properties of an effective conditioned stimulus and would elicit this leg flexion response. So, the two groups were trained in this task 

In both repeated images, another arrow is drawn from the blue circle to the red circle, and is labelled ‘contiguity’.

[Jose] where there is contiguity between the light and the shock, consistently the light was always followed by the presentation of the shock. But, in one of the groups, the animals were made familiar with the light before the onset of the training in the conditioning task. So, there were around 20 or 40 presentations of the light before the actual training

A graph is added to the right side of the slide showing two increasing lines, one white and one black.

[Jose] and the exposure to the to be conditioned stimulus resulted in a deficit in learning. Here we can see this is the performance of the group given pre-exposure that was familiar with the light and this is the control group. This is the normal, the baseline acquisition of a conditioned response and this difference, this deficit, is what latent inhibition is about. So, in spite of the contiguity between the light and the shock, the association was hindered. Why does this happen?

A black box appears at the bottom of the slide with white text inside that reads:

During pre-exposure, attention to the light seems to decrease and the light loses capacity to acquire predictive value. To account for latent inhibition we need to appeal to selective attention processes.

[Jose] Well, something happened during the pre-exposure, the animal learned something. Apparently, the attention paid to the light went down, and at the time of conditioning, the animal was not paying attention to the light, which loses then the capacity to enter into new associations. So, to account for that inhibition contiguity is not enough, we need to refer to selective attention processes and this would be one of the characteristics of the unlimited associative learning.

The slide changes. Black text on the slide reads:

CS-US contiguity is not a sufficient condition for learning - II

Sensory Preconditioning (e.g., Brogden, 1939; Rizley & Rescorla, 1972)

A diagram is shown below the text, showing three sections – Phase one, Phase two and TEST.

[Jose] Another example, which is even better than this, is sensory preconditioning. Sensory preconditioning is the twin brother or sister of second order conditioning, and this is an experiment in three phases. And again, what we are doing is simply training two groups of animals in a standard Pavlovian task where a light becomes the signal for a shock. This experiment was done by Brogden in the late 30s with dogs and by Rizley and Rescorla in the early 70s with rats.

Before training in the conditioning task, one of the groups had some experience with the to be conditioned stimulus and the third stimulus in that phase was a tone. So, in the case of Brogden, a tone and a light were presented together simultaneously as a compound and in the case of Rizley Rescorla, the tone preceded the presentation of the light over a number of trials. A control group would be given exposure to the tone and the light but at random, there was no way in which the animals could establish a relationship between these two stimuli because there was no relationship. And, the final test was done with the tone alone. What they wanted to know was whether this experience in phase 1 would affect how the animals behave in front of this tone stimulus.

So, what can happen in this kind of experiment? In the first phase, the tone might become a signal for the light and in the second phase in both groups the light could become a signal of the unconditioned stimulus which would become an effective conditioned stimulus and would result in conditioned responses, specifically fear conditioned responses. It is important to note that in this design, there is no contiguity between the tone and the shock. The tone and the shock have never been presented together and furthermore, when the tone was experienced with the light, the light was not yet an effective conditioned stimulus. So, there is no way the tone could be linked with the fear response at all during the first phase of the experiment.

In spite of all these limitations and the lack of contiguity, when they tested the animals the animals in the group given pairings of tone and light and light and shock, showed an effective fear conditioned response. So, the tone was an effective conditioned stimulus in spite of the fact that it had never been paired with the unconditional stimulus, there was no contiguity at all. The only way in which we can account for that result is by assuming that the animal established different associations, acquired different information in the two learning episodes, learned first to link the tone with the light and then associated the light with the unconditioned stimulus. These associative triggers would allow the tone to be an effective, conditioned stimulus.

A black text box is added to the bottom of the slide with white text that reads:

The animals are able to integrate information from separate learning episodes

[Jose] So, to account for the sensory precondition effect we need to assume that animals are able to integrate information from independent learning episodes. The animals learn things at different times than then can be linked together and this changes the way in which they behave in front, in that case of the term. Impressive, isn’t it?

The slide changes. Black text on the slide reads:

Five key features that define Unlimited AL

Unlimited AL goes beyond the mere learning based on contiguity. It requires:

Ginsburg & Jablonka (2019); Birch, Ginsburg & Jablonka (2021)

[Jose] So, this is the most difficult one, I think. Ginsburg and Jablonka Came up with five key features that define unlimited associative learning. To be honest with you, of the five I have selected four because the fifth I couldn't understand and I have replaced that one by myself, which is in green. But basically, what they are saying is that to be considered unlimited associative learning the animals must be able to use selective attention, integrate information from different learning episodes and show flexibility, so one conditioned stimulus can change value if the conditions in the environment have changed, so something can be indicative of something repetitive one moment and then this can change if the conditions in the surroundings have changed as well 

It needs to be complex. The hierarchical structure or the associative structure behind learning can be hierarchical in such a way that one stimulus can become, for example, a retrieval cue for the association between order to a stimulus or the context, for example, can activate the association between two stimuli. And finally, the animals must be able to distinguish the compound form elements, so A and B together are not the same as A and B presented by their own. This defines unlimited associative learning and requires some level of cognitive sophistication.

The slide changes. Multiple text boxes appear creating a three by four table. The headings for the table are ‘Cognitive Complexity’, ‘Phenomena’ and ‘Mechanisms’. Down the left side, the table is split into two - ‘Consciousness’ and ‘Sentience’. Under each heading are multiple methods and strategy names.

[Jose] So, with this in mind, then I perpetrated that. This is my contribution; I have captured the five conditions for unlimited associative learning here and I have identified a number of basic learning phenomena (Pavlovian phenomena) that would require the usage of these skills. So, this goes from very simple learning on associative sensitisation and short-term habituation to limited associative learning and unlimited associative learning.

This is where my wisdom reflects superior cognition. Superior cognition would be memory, the capacity to learn from the past, plan for the future, theory of mind in being able to read the minds of others and be aware of the intentions and feelings of others to use it and having the capacity to interact with the environment by using things which are external to your own body.

All of this may be superior cognition that depends on mechanisms that I have labelled as “complex systemic brain activity”. Meaning, I don’t know. This is for my colleagues to fill, the cognitive neuroscientist in the room might have some ideas.

So basically, what we want to do is to assess which animals qualify for unlimited associative learning

To the right of the table, a green text box appears with white text that reads ‘Vertebrates’. The image of the historic man is also added.

[Jose] and I would say that humans and vertebrates in general are typically assumed to be able to show unlimited associative learning. We have seen that dogs and rats are capable of integrating information from different sources etc.

On the bottom right, a cream text box appears with black text that reads ‘Other Kingdoms’.

[Jose] We also know that organisms from other kingdoms such as Protista or plants and fungi are able to learn, they learn from experience. Their behaviour can be changed as a consequence of experience, but typically, we assume that they would show very low-level learning.

In the middle of these two text boxes, a blue text box appears with white text that reads ‘Other Animal Phyla’.

[Jose] And then somewhere in the middle, we will have the phyla; These are referred to in general in as invertebrates. So, there is the molluscs, the arthropods, the nematodes, the earthworms, etc. So, these other animals, phyla, might sit here according to some. Or, according to others, they might be as competent as the vertebrates and would show unlimited associative learning. I suppose that they sit there in the centre. Our task is to assess whether different species representing the animal phyla actually meet the criteria for unlimited associative learning.

The slide changes. Black text at the top reads:

Tracing the emergence and evolution of learning

Below this is a circular diagram, with multiple lines and boxes inside. Around it, there are labels – Plants, Animals. Fungi, Bacteria, Archaea and Protists 

[Jose] This is the phylogenetic tree presented in a circular fashion. I like it because this shows us where we are, which is here. This shows the Homo sapiens sitting amongst thousands of species in the animal kingdom and millions of species in general, so this is just to humble ourselves. We are just one species amongst many, many millions.

The plan for today is to review some evidence about the learning abilities of different organisms and I'll start very basically with Protista. They are single cells; complex cells in many cases but yes, one single cell. I'll then review two experiments with plants, which are of some merit and then I will jump from the plants to animals with a brain and a centralised nervous system (I am skipping the animals with a central nervous system but no brain because there is really not much evidence about them. They have not really been established).

The slide changes. In the top left corner, a red text box with white text reads ‘Limited AL’. Next to this is a purple text box with white text that reads ‘Simple Conditioning … contiguity’

Underneath is a grey diagram showing a Capillar Tube. This diagram is labelled ‘Exp 1: Learning in Paramecium caudatum’

[Jose] Starting with the Protista, There is a long tradition that goes back to the early 20th century, suggesting that single cells like their Paramecium Caudatum can learn and many important people at the time even claimed that these single cells behaved with purpose and have conscience. I think that this is a very, very long a stretch, but they certainly have some learning capacities 

So, this is an experiment that was published by Hennessey, Rucker and McDiramid in 1979 and I favour this one because this was published in Animal Learning and Behaviour and for some reason I tend to trust what is published in that particular journal. So, this is a capillary tube and the authors of this study simply captured one single cell in this capillar tube so that they could pass an electric shock through the medium. The paradigm that they used was a Pavlovian conditioning paradigm and they simply presented to be conditioned stimulus which was a vibration that was produced by a little speaker close to the capilar tube. This is done under a microscope and following the presentation of the to be conditioned stimulus they presented a shock and the presentation of the shock produced a conditioned response of spinning.

So, they repeated that a number of times and by pairing the vibration (the conditioned stimulus) with the shock a number of times they found that the vibration by itself produced a spinning action. They ran several experiments; I'm going to tell you about two of them.

An orange text box appears with black text that describes what happens in this experiment, such as ‘Conditioning’ and ‘Extinction’, ‘Vibration’ and ‘Shock’. An orange line graph also appears on the right to show the results from this experiment.

[Jose] This is the first experiment where they trained two groups of cells and one of them was given the standard Pavlovian training where the vibration was followed by the shock in every trial, and the second group was given random presentations of the vibration and the shock. What they found was that those cells that were given paired presentations of the vibration and the shock actually developed a good level of conditioned response. This is the probability of showing that they're actually spinning in the presence of the vibration alone. And, as you can see, there is some level of response in the control condition, but it is not as high as the one in the group given the pairings between the shock and the stimulus. So, this difference here seems to be good evidence for Pavlovian conditioning in single cells.

In a subsequent phase of the experiment, they omitted the unconditioned stimulus of the shock and they saw that after a number of trials, the conditioned response extinguished. So, this is a basic characteristic of any Pavlovian conditioned response. It looks like single cells have gotten capable of learning the relationship between two events in a Pavlovian preparation.

A grey text box appears underneath the orange box, with black text that reads ‘Long Term Memory’, with columns labelled ‘Day One’ and ‘Day Two’. To the right of the orange graph, a grey graph appears showing the results from day two of the experiment.

[Jose] In the second experiment, they presented one group the vibration followed by the shock and then they waited 24 hours before continuing with their training alongside two more groups that were given the vibration with the shock paired or vibration and shock at random. What they wanted to know was whether a single complex cell, would be able to retain the information acquired in day one to facilitate the acquisition in the second day of the experiment. So, this is a memory experiment. What they found is that the group that was pre-trained in day one actually showed a great advantage. So, the cells (not the animals, not the plant) actually retain the information for 24 hours. This seems to demonstrate that not only can single cells learn a Pavlovian association, but they can also retain the information long term, this is an instance of long-term memory.

The slide changes but the red and purple text boxes remain at the top of the slide. A dark red text box appears with white text that reads:

Long-term predictive learning can take place in the absence of a nervous system!

[Jose] So how does this happen? I was very sceptical of these kinds of things initially but there is a number of classic experiments conducted in the fifties and some more recent experiments that have been replicating the classical studies and there is evidence piling up, suggesting that individual cells can actually show Pavlovian conditioning. So this is happening, they can show long term learning and memory. The question is how?

A dark blue text box appears with white text that reads:

Long-term learning and memory depend on an epigenetic process, the synthesis of proteins, in the nervous system of neural animals (Davis & Squire, 1984) as well as in complex cells (Ginsburg & Jablonka (2009).

To the right of this box is a diagram that shows Protein Synthesis.

[Jose] We need to look again at the work of Ginsburg and Jablonka and they have suggested that learning requires an epigenetic process and this is common to neuro animals and single cells. So, what we understand nowadays is that when they are exposed to the learning experience, to the events which are part of the Pavlovian paradigm, this switches on one gene which then triggers the transcription process and the result of that is a messenger RNA, which is transported out of the nucleus and the ribosome reads the code and produces a new protein. So, the synthesis of this protein is the product of the learning episode, and this new protein actually changes the structure of the cell which will now react to the conditioned stimulus in a different way than they reacted initially, so they would hopefully react in a more adaptive way.

This is common to the single cells and the neural animals. Protein synthesis is necessary for long-term encoding of information. If you treat the animals (rats for example) with an inhibitor of the protein synthesis they fail to remember the information acquired in a learning episode.

Black text appears underneath the text boxes, and reads:

Learning appeared very early in evolution; the basic mechanism has been maintained and is currently in use in specialised tissues that process information.

[Jose] So, this suggests that the emergence of learning was present very early in the stages of evolution. It's a very old mechanism that was already present in individual cells and this might have been maintained throughout evolution and is now in use in specialised tissues that process information (that would be the nervous system of the animals).

The slide changes with the red and purple text boxes remaining in the top left. However, purple box text now reads ‘Long-Term Habituation’. There is a black title on the page that reads:

‘Exp 2: Learning in Mimosa pudica’

Underneath is an image of the Mimosa pudica plant with someone touching it. Black text below this reads:

Mimosa pudica folds in response to mechanical stimulation.

[Jose] So, let's jump from single cells to a complex organism, plants. Do plants learn? Plants are really difficult to deal with, they don't really do anything except for a few species which actually react immediately to stimulation. Mimosa Pudica is one of those species that present a very clear response to mechanical, electrical or chemical stimulation. When you stroke one of the pinnae of the Mimosa Pudica (the little branches are referred to as a pinnae) they fold immediately and this immediate response has been used to assess whether plants can learn or not.

Another image appears on the right that shows an article titled ‘Experience teaches plants to learn faster and forget slower in environments where it matters’. A diagram of a plant being measured is also shown next to this.

[Jose] I was a sceptic, as you can imagine but in 2014, Monica Gagliano came up with a paper in which she demonstrated long term habituation in plants. It was beautifully written, very convincing and very persuasive but I was still not buying it.

A new article image appears underneath this, with the title ‘Leaf-folding response of a sensitive plant shows context-dependent behavioural plasticity’. A diagram of a plant is shown next to this, with a red circle outlining a leaf.

[Jose] Simultaneously, in 2014, Sabrina Amador-Vargas came up with another paper with Mimosa Pudica again, demonstrating habituation learning. The two studies differed a lot; Monica Gagliano was stimulating the whole plant. Basically, what they did was to drop the plant pot from a height of 15 centimetres and then all the pinnae would close. Sabrina Amador-Vargas was more subtle, and she was stimulating just one leaf of the pinnae. They dropped a grain of rice on one leaf, and then they measured the folding of the individual leaf. In some cases, I lack the expertise to deal with plants to be honest with you, So I was struggling to believe that this actually happened.

A new image appears in the right corner of the slide showing a diagram of the plant experiment designed by Professor Jose Prados.

[Jose] So, I decided to do some experiments with plants myself and rather than doing the brutal stimulation of the whole plant or the subtle stimulation of one leaf, we decided to stimulate one pinnae and see how the pinnae folded.

So, to do experiments with plants, you need to be well prepared. You need a very rigorous protocol,

The slide changes. A purple text box is in the top left corner, with white text that reads:

Plant Learning: Preliminary Work Protocol

Below this is a cycle of numbers ranging from one to seven with arrows pointing to the next consecutive number. As Professor Jose explains this cycle, images will appear where the numbers are, showing the steps of the experiment. Professor Jose completed this experiment with the help of a young boy who is also pictured in the images.

[Jose] and in my case, the protocol took around seven steps. The first thing is that you cannot do it alone, you need to do it with a research assistant which needs to be motivated and needs to have the right expertise. I was lucky enough to get this guy here who helped me in planting the seeds. And then, over days and weeks, we nurtured the seeds. We respected the conventions about temperature and moisture and the seeds germinated. We then transplanted them and engaged in mass production because I had the plan of running a complex experiment with different groups and different variables, I was ambitious at this stage. When the plants were sufficiently grown, then you could do two things. One was to experiment, which is the boring one. The other one was to bring them to school for show and tell. I wonder what happened to this plant with 20 boys poking at them over one day? *laughter*

The slide changes. At the top of the slide, black text reads:

We can measure the folding and the reopening

In the top left corner is an image of a pinnae branch, with an arrow across it measuring. Black text next to this reads:

Time: 0
Before stimulation d max = 12mm

[Jose] So, this is what I did. I brought the plants to the lab and we wanted to measure the folding response and the reopening response so how do you do that? Before stimulating the pinnae (we did that with a little brush) we measured the distance between two of the leaves and this was giving us the maximum distance 

A new image appears underneath showing the pinnae closing. Black text next to this reads:

Time: 5 sec
After stimulation d min = 1mm 

A purple text box appears under this, with white text that reads an equation for the percentage of d max.

[Jose] So, after a few seconds, what you'll get is the minimum distance once the pinnae have folded and if it is one millimetre then you get a percentage of the maximum distance, which is in that case 8%. So, 0% would be the total closure, 75% would be a weak closure and 100% would be an absence of folding.

Two new images appear underneath, the first showing the pinnae reopening. Black text next to this reads:

Time: 10 min
After stimulation d n = 6mm 

A purple text box appears under this, with white text that reads an equation for the percentage of d max.

The second image is showing the pinnae fully reopening with black text next to this that reads:

Time: 20 min
After stimulation d max = 12mm 

A purple text box appears under this, with white text that reads an equation for the percentage of d max.

[Jose] Then, you can monitor the reopening, so after around 10 minutes, if it is, for example, six, you have already had  50% and after about 20 minutes we were expecting the plant to be back to normal at 100%.

Something I learned is that if you are not an expert in plants, then you cannot do big things. The simpler the experiment, the better. We did several things with different students, we tried to use electrodes and it was a complete disaster. But, when the manipulation was really, really simple, we got some decent data. With Jack Hugues, we did this experiment in which we simply trained the six plants on a habituation task. So, we simply stimulated the pinnae 8 times within a session and our expectation was that the first time they would show a strong fold in response and a slow reopening, and then in trial, 6, 7 and 8 of the same session they would habituate and the folding would be weaker and the reopening would be faster. But, what we found was the opposite within our session. What happened was that the pinnae closed to the maximum closure possible and then the reopening was sower and slower. So, Jack was expecting to spend like three hours in the lab, but he spent the whole day at around 6, 7, 8, or 9 hours, because the plant didn't reopen at the pace that we expected.

Next to the column of images, a grey graph appears with the black title:

Hugues and Prados (2016): Experiment with 6 plants
Habituation training (4 Days x 8 trials)

In the graph, there is a line for each of the 4 days, showing an increase in what happened over time during that experiment.

[Jose] So, after the first day, having to spend about 8 hours in the lab I managed to persuade Jack to return on day 2 and repeat the operation. After the second day he thought it was a complete disaster because we were looking for habituation and we got sensitisation and so I couldn't convince him to come back the third day on Wednesday. He said he had an excuse. “I have a deadline for an essay”, so I say, “OK, when is the deadline? Thursday? come back on Friday”. A bit of negotiation later and he came back on Friday and then after another long session on Friday somehow I persuaded him to come back on Monday and repeat this nightmare of an experiment for the fourth time. And the reason to be so persuasive was that I had a cunning plan. I had realised something between day one and day two and it was that rather than looking at the evolution of the behaviour within the sessions, you should look at the evolution of the behaviour throughout the sessions, comparing throughout the sessions and especially looking at the first daily trial 

So, this is what we observed the first time that we stimulated the Pinnae. There was a closure of about 12% which is quite big and then the reopening took this shape. It took around half an hour for the pinnae to reopen in full. The second day we got this, which is a similar folding response but with a faster reopening and this faster reopening would be indicative of habituation, but not only habituation, it’s long-term habituation. The experience of the training session 24 hours before was increasing the speed of reopening and on the first trial of the second session. Then, in the third session, what we observed in the first trial was this, something similar. This is good because it shows that in the second day you begin to artefact, so this was for real. And this is the fourth session, so what we got was a significant effect of reopening compared to the first day and days 2, 3 and 4 in the first trial, and on top of that, on the 4th day we got a weaker folding response.

A black text box appears underneath the graph, with white text that reads:

Plants show long-term habituation, how?

[Jose] By itself, this would be encouraging the evidence found in those two papers. I was now convinced that plants actually learn and show long term habituation. How do they do that? I don't know how, it's beyond my understanding. For the short-term habituation ionic channels have been suggested, but for these plants to retain over 24 hours, there must be a structural change in the plant that accounts for that and I have no idea how this works. This is for the plant experts to decide.

The slide changes. Red and purple boxes are in the top left corner. Below this is an image of an article from ‘Scientific Reports’ with the title ‘Exp 3: Learning by Association in Plants’

[Jose] So, I thought that I had some cool results with the long term-habituation and then Monica Gagliano came up with this. This was published in Scientific Reports, and it's titled “learning by association in plants”. So, she trained plants in a Pavlovian conditioning task, and she got actual Pavlovian conditioning, which is mind blowing. So, what did she do?

Three blue boxes appear on the slide, each with a diagram explaining how Monica Gagliano completed this experiment.

[Jose]They exploited the fact that plants orient towards light, light is very important to them so light is the unconditioned stimulus. They found that if a plant is put in a Y maze and a blue light is presented in one of the arms. Subsequently, when the light is off, the plant tends to orient forwards the arm where the light was presented, and this is considered an unconditioned response. So, in the proper experiment what they did was to present together two events, the blue light and an air current, which was produced by a fan, and the rationale was that the air current would become a signal for the blue light, and they expected this effective conditioned stimulus to attract the plant, even if it was presented in a place where the light had never been presented before. So, in the test, what they did was to present the air current in the opposite arm and the unconditioned response would be to go to the left but the plant orientated for the air curtain. So the air current was working as an effective conditioned stimulus, as a signal for the light.

This is an example of a conditioned response. And if this is cool, wait a second and see because the last experiment was amazing. In that case, they presented the air current and the light in the opposite sides of the maze in such a way that if anything, the air current could become an inhibitor, a signal for the absence of light. Then they presented the air current in the arm where the blue light had been presented and in spite of that, the plant decided to grow in the opposite direction to the air current. So, the air current was actually a condition inhibitor and what they demonstrated, according to that, was an inhibitory condition response.

A black text box appears in the middle of the slide with white text that reads:

Plants show associative, causal learning.
Do they? If so...how?

[Jose] So, plants show associative learning. To be honest with you, I lack the expertise to properly assess these experiments. They are really persuasive, and I have to buy it. I know that some colleagues in plant learning labs are interested in replicating this and I would like to wait until this is replicated independently by other researchers before concluding that causal learning can be observed in plants. As for the how they do it, your guess is as good as mine. Honestly, I have no idea. 

The slide changes. In the top left corner, a yellow box with black text reads ‘Unlimited AL’ and a blue box with white and yellow text reads ‘Blocking. Prediction error/selective attention’.

Underneath this, black text reads:

Experiment 4: Learning in Planarians

Below, there is an image of three Planarians on a white ground.

[Jose] I'm jumping now from plants to animals, and these are animals with a proper brain and a centralised nervous system. I will start at the bottom of the scale with planarians. Planarians are members of the phylum of Platyhelminthes, and this is the phylum which is farthest away from vertebrates in the phylogenetic tree but that still retain the same brain and nervous system. There are similarities between the nervous system of the Planarian and the vertebrates, including mammals. So, the neurons are quite similar and they express the same neurotransmitters and (according to our research in the lab) with the same function. So, the dopamine antagonist has the same effect in the planaria in a conditioned response task as it has in rats and so planaria are suitable for learning and studies.

The slide changes, but the yellow and blue text boxes remain in the top left corner. Below is a blue box with a diagram showing blocking and control with light and vibration. A black box below this reads:

To the right is an image of a paper titled ‘Cue competition effects in the planarian’ written by Professor Jose Prados and many others.

Further down the slide, on the left, are two white graphs showing the results of tests A and B. A light blue box on the right reads:

Challenges the principle of contiguity.

To account for blocking, we have to refer to:

[Jose] The question was whether planaria would be able to show evidence for limited associative learning. When you stimulate the planaria mechanically or electrically, they shrink in response. So, if you shock them or you touch them they immediately shrink and we can use that as an unconditioned stimulus. In the experiments that we ran with the planaria we presented a conditioned stimulus using light or a vibration counterbalance. In principle, the light and vibration were of low intensity and didn't produce any apparent response and the unconditioned stimulus was a shock. 

In this kind of experiment (and in this one in particular) we train the animals in the Pavlovian task, this time with a compound stimulus; a light and vibration were presented together as a signal for the shock. The animals shrink in the presence of the shock and then during that we presented one of the elements to see whether in the present of B, they show the conditioned response of shrinking. One of the groups, however, had previous experience with the other element, so in the blocking group, the A element is pre-trained as a signal for the shock, and we know from the broader literature that there are thousands of demonstrations of the blocking effect that when you do that in the blocking groups of train A and train A and B together, B doesn't acquire the properties of a conditioned stimulus. They fail to learn about B.

So, A is a signal of the unconditioned stimulus. When you present A and B, A is already signal of the plus and B is redundant and somehow the animals filter out this B element which doesn't become associated with a shock. The question was whether we could replicate this experiment in planaria and the results simply show that in the first two phases of the experiment all the animals in the control and the blocking condition acquired the conditioned response as suspected, and when we tested B, we observed a deficit, the characteristic deficit of the blocking effect in the experimental group. So, we replicated the blocking effect in planarias, and this is big. It took quite a lot of stamina to complete this experiment.

This by itself challenges the principle of contiguity. The outcome continues during their training. So, there is something here that is more important than mere contiguity, and this extra thing could be the predictive value of A which is carried from the first phase of the experiment to the second. Or it could be selective attention; the animals filter out B, which doesn't have access to the learning mechanism and there is no opportunity for it to become associated with the unconditioned stimulus. So, flatworms seem to qualify for unlimited associative learning. I'm highlighting that in yellow because I would like to be cautious, and before we claim that they are capable of unlimited associative learning and conscious state I would like to see other phenomena demonstrated within that flatworm.

The next slide appears. In the top left corner, a green box with white text reads ‘Unlimited AL’ and a blue box with white text reads ’Sensory Preconditioning. Integration of information’.

The black title in the middle reads:

Experiment 5: Learning in Snails

Below this are three blue boxes with diagrams of snails inside. The first one shows ‘Feeding Behaviour’, the second shows ‘Training’ and the last one shows ‘Test’.

[Jose] Experiment number five is done with sails. Just to show you, snails are one member of the mollusc family, gastropods. And the kind of experiments that have been done with snails involve snails eating. The key aspect is that when they are eating, the tentacles are lowered, and this is a natural reaction which appears when they are feeding. So, someone came up with the idea that you could present some fruit simultaneously with a feeling and this would be equivalent to the A-plus treatment. During the eating of course, the animals lower the tentacles and then when you test the other by itself, if the animal lowers the tentacle, then that can be considered a conditioned response.

The slide changes, but the previous green and blue boxes remain in the top left corner. An image of an article is added with a black title reading:

Conditioning of tentacle lowering in the snail (Helix aspersa): Acquisition, latent inhibition, overshadowing, second-order conditioning, and sensory preconditioning
By Ignacio Loy and others 

Below this, the three phases of experiment 5 are shown for two groups, Group COM and Group ELE. To the right of this, graphs appear showing the results of this experiment, with the Apple and the Pear.

A black text box then appears with white text that reads:

Snails are able to integrate information from separate learning episodes, a key feature of Unlimited Associative Learning

[Jose] So, using this procedure, my friend Ignacio Loy has done many experiments, and I want to highlight this one on sensory preconditioning. Remember that sensory preconditioning is the one done with dogs and with rats in which animals could integrate information coming from different learning episodes. So, he and his team replicated that in an experiment in which animals were given Pavlovian conditioning with one other followed by the presentation of food (in this case apples and pears). This is supposed to imbue the apple with the properties of an effective, conditioned stimulus. But before that, the animals in the group were given presentations of the apple and pear others. These were presented simultaneously in such a way that the animals could associate them.

The control group, group element or Elemental, was given access to apple and pear separately in such a way that they couldn't associate them. In phase 3, They presented the pear and they wanted to know whether they would show these the conditioned response in the same way as dogs on rats do. The results show that the animals acquired the conditioned response to the apple and when they were tested with the pear, the group given compound presentations of the apple and pear showed a very strong conditioned response to that pear, in spite of the fact that this stimulus had never been paired with the food itself. So, snails actually are capable of integrating information from different learning episodes and this is really quite sophisticated. This seems to suggest that snails meet the criteria for unlimited associative learning, and this might suggest that they are capable of experiencing conscious states. Last one...

The slide changes but the green and blue text boxes remain. However, the text in the blue box has now changed to ‘Positive and Negative Patternings. Discrimination of pattern from elements’. Below this is a black title that reads:

Experiment 6: Learning in Honeybees

Two images of honeybees appear underneath this, showing the honeybees in the experiment ‘Conditioning of Proboscis Extension Reflex (PER)’.

[Jose] The last experiment was with honeybees. Our colleagues that work with honeybees do that by restraining them in a harness and then they present an odour which is a conditioned stimulus and then a conditioned stimulus and unconditioned stimulus which is sugar water. The animals in the presence of the sugary water extend their proboscis, and when you pair the odour with the sugar water for a number of trials the extension of the proboscis can be considered a conditioned response.

The next slide appears, and the text boxes in the left corner remain. Below this, another article image is added. This is titled:

Configural Olafactory Learning in Honeybees: Negative and Positive Patterning Discrimination

By Nina Deisig and others 

Underneath are two diagrams. One showing the ‘Negative Patterning’ process and the other showing the ‘Positive Patterning’ process – using an image of a bee. To the right of these are two graphs showing the results from both tests.

A black text box underneath has white text that reads:

Honeybees discriminate the elements from the configuration, a key feature of Unlimited Associative Learning.

[Jose] So, using this procedure, some colleagues assessed whether honeybees can actually discriminate a component from its elements which requires quite a high level of sophisticated cognition. So, they trained the animals in a negative patterning discrimination where the elements, the other 1 and the other 2 were trained as signals for the unconditioned stimulus for the sugar water, so the animals extended their proboscis. Then, they interspersed trials in which the two others were presented together, but when the two were together the unconditioned stimulus was omitted and if the animals learned this complex discrimination, they would respond to A and B, but they will inhibit the response to the component of A and B, showing that they can distinguish between the compound and the elements.

The result of the experiment is summarised here. The animals responded to A and B and inhibited the response to the compound. This is quite good. And then, to round the result, they also trained the animals in the positive patterning, where the other 1 and 2 were paired with the absence of their unconditioned stimulus and only when the compound was presented, the unconditioned stimulus was presented. In that case of course, the animals responded to the component but didn't respond to the elements. this is brilliant and shows that honeybees can discriminate between a complex pattern and the element, which is, according to Ginsberg and Jablonka, a key feature of unlimited associative learning.

The slide changes to show a slide used previously. Multiple text boxes appear creating a three by four table. The headings for the table are ‘Cognitive Complexity’, ‘Phenomena’ and ‘Mechanisms’. Down the left side, the table is split into two - ‘Consciousness’ and ‘Sentience’. Under each heading are multiple methods and strategy names.

To the right of this, a green text box appears in line with the consciousness row, with white text that reads ‘Vertebrates and Arthropods, Molluscs, Flatworms...’. Below a grey box appears with white text that reads ‘Nematodes, Annelids, Echinoderms, Rotifera, Cnidaria, Ctenophora’. At the bottom, and in line with the sentience row, a yellow text box appears with black text that reads ‘Other Kingdoms’.

[Jose] So, to finish. We tend to agree that vertebrates and capable of unlimited associative learning, and now we can add arthropods and molluscs to this list of species that can be eligible for having conscious states. I would like to add the flatworms, but as I said, I would like to have some more evidence. We've learnt that other kingdoms, plants and single cell organisms, can actually display associative learning according to the literature and that there are a number of animals for which we don't have enough evidence for. So, we don't know really whether it's there.

The slide changes to have black text that reads:

Conclusion

[Jose] So, as a conclusion, there seems to be evidence that different animal Phyla can display unlimited associative learning and are therefore eligible to be considered conscious, and this includes the vertebrates, arthropods and molluscs and perhaps also the flat worms. Many species have not been properly assessed and in the coming years we should be able to produce evidence to complete the picture, at least by selecting some species representative of the different animal phyla. We cannot assess all the animal species because there are millions of them, so one per phyla would be better than nothing.

This behavioural evidence can be considered alongside other sources of information, anatomical neurochemical, neurophysiological etc to establish whether animals meet the criteria to be considered sentient, to have feelings or even be capable of conscious states.

The next slide appears. This has black text that reads:

Our Next Steps

Underneath this is an image of two computers with images of the experiments on. This is labelled ‘Planaria Lab at Derby University’.

[Jose] This is the planaria lab at Derby, we managed to bring it from Leicester, and this is in a secret location like the Bat cave. Seven animals have been recorded and the next step would be to get some funds to run more of these experiments in review based on this talk and trying to observe secondary conditioning and continue using these animals as models for addiction and anxiety related disorders. This is something that we've been doing over the last 5, 7, 8 years with Gonzalo Salai and several PHD students. The last thing would be to perhaps, being sentient creatures, we could use the planaria to develop models for pain and pain management which is one of the great research areas in psychology and the college in this University.

The slide changes and black text appears that reads:

And just for the record...

[Jose] And, for the record, last Sunday in The Guardian, they addressed the decision by the RHS, which stands for the Royal Horticultural Society not to consider gastropods and weeds as pests anymore. So, to all the ones that were killing snails in your gardens, shame on you and on me. They have discovered that killing them is more harmful than beneficial. They are part of the food chain and they have discovered, for example, that weeds are essential because they are pollinators. The local pollinators prefer the local weeds over their alien plants that we insist on growing in our gardens. So, from now on, stop killing the snails and stop removing the weeds. Something to think about.

The research we are doing might contribute to raise awareness about the relatively advanced cognition of animals in general, including invertebrates. This may help us to better understand what is their place in nature, and if we do that, I'm sure that this will help us understand what our place in nature that would contribute to a more respectful and protective approach to our environment. We would be better people, and this would increase our health and wellbeing. So don't kill the snails.

The slide changes and black text in the middle reads:

The last couple of slides suggest ways in which this research can fit within the research strategy of the University of Derby.

[Jose] And finally, if you were thinking, “what the heck is all that about?” I think that those last couple of slides might suggest ways in which this kind of research can fit within the research strategy of this university. Thank you very much.

The final slide appears. An image overlooking Derby fills the whole slide, with white text in the middle that reads ‘Thank you’. The University of Derby hills logo is in white and in the top left corner of the slide. In the right corner, in white, are the University of Derby contact details and social media account links.

(Applause)

[Paul] Thank you very much indeed Jose and thank you for the strong environmental message at the end. A wilder Derbyshire should be better for all of us. The frogs in my garden love the snails!

As the university botanist I never thought I would be coming to listen to someone talk about plants, so thank you for that. I'd really like to see those papers. Any questions for Jose?

[Audience member] In terms of from an evolutionary, biological point of view, if we allow evolution to run its course without any kind of extinction event are we saying that all the phyla would make some kind of evolutionary progress towards sentience? Unlimited learning, consciousness and beyond?

[Jose] Good question. It’s the planet of the apes question. I tend to think, yes, if we don't kill them first. And the common agreement is that life tends towards complexity. So, to our knowledge, there was only one biogenesis, and this happened in our planet once, so there were not several biogenesis’s.

So, I've heard many times that where there is life, you can expect complexity to happen. And if there is a biogenesis on a different planet from Earth, we can expect it to tend towards complexity and generate civilisation. So, in principle, I would say that say that if it stays alive for several millions of years then Yes, they will eventually become sentient and conscious.

[Kathryn Mitchel] I suppose my question is more about the last statement in terms of how we will use this knowledge. I suppose what I would want to know is where is that research going to be applied to things like anxiety and particularly pain, because Pain's another key area. So, I’m just interested because I think we do need to look at that because I think it would be an interesting area to develop further and one that I think we have the beginnings of and how would we then invest?

[Jose] I think that it's an interesting thing. I've been discussing it with several people in the community and there is some interest, and I think that it might fit within the research strategies. But whether it will happen or not I don’t know.

[Kathryn Mitchel] I suppose it can happen if we plan it to happen. I spend a lot of time now with the trust and I think we need to plan it because I think it does fit with quite a lot of the other areas and I suppose I'm interested in how you think that would work?

[Jose] What I would like to avoid is becoming a collector of learning phenomena. I like to be more strategic and use the knowledge that we have acquired playing with the Model and to put it at the service of the practitioners. So, developing models for addiction or anxiety related disorders or pain management is something that has a future. If I continue doing research on inhibition, blocking and several other conditioning processes, I will retire with all the glory (laughs)

[Audience Member] That was a fabulous talk, I really enjoyed it. I'm not a psychologist by any stretch of the imagination, nor a botanist so my question is just a naive question in a way. With the leaf folding and the response and how it's interpreted as learning could it be simply that the stimuli and the very mechanical presentation on the Leaf is causing an increase in receptors within the leaf, and therefore it's causing over stimulation? And in any sort of system, if you do too much of anything, it’ll go to the extreme as opposed to actually learning so was there any discussion in the papers with respect to that?

That’s first question, the second question was in the experiments with the fan and the light, does the fan effects the carbon dioxide level or do they actually think about that either?

[Jose] So, this is about plants, and I have no idea about plants to be honest. So, it beats me. I don't really know how it how it happens but what I can tell you as a fun fact is that I didn't show you the whole data that I got. I was interested in knowing whether these changes in the behaviour of the plant when we were doing habituation is local or if it affects the plant as a whole. And so, on the last day of training, at the end of everything, I asked Jack to stimulate a different pinna of the same plant and to our surprise, we found that it showed a lower level of folding and more rapid reopening than when we first stimulated the target pinnae, suggesting that this is a systemic change that we are recording. But honestly, in the absence of controls and preliminary work and continuing with that for a few years, I'm not in a position to say anything sensible.

And the second question was about the Gagliano conditioning experiment. The air current was supposed to be a neutral stimulus, and they assessed that, but I don't really know because the paper is written in such a way that it is really, really convincing but sometimes you don’t get the whole level of the detail in the procedures by this lab. So, I'm not in a position to critically assess the procedures. But I'm very interested in knowing whether other labs would be able to replicate that, and I know that they are looking into it. Probably in a year’s time we could have another gathering and I can tell you about the results.

[Warren Manning] First of all, fantastic presentation, I really enjoyed it and as somebody who doesn’t understand this world, I picked up quite a lot. I'm interested in those five conditions; does it matter if you went to the world of inorganic materials, manufactured materials and AI that create the same responses and then does that raise the question then whether those five conditions are true or not?

[Jose] I forgot to mention that. I was driving today point and thinking “I need to mention artificial intelligence.” I think that these fundamental markets for unlimited assistive learning are of importance because then there are systems, artificial systems now that show learning and that control our lives and at some point, we will have to decide whether they are capable of conscious states are not. And I think that we need to get ourselves some tools to assess the question of whether they are experiencing consciousness states.

[Warren] I was also thinking that it was really interesting with the work of plants and before that, the work on what was happening within the cell that was exhibiting similar behaviour and I started to think “My consciousness has always been there, it's the network that creates it”. We have things like shape memory alloys and engineering that exhibit exactly the same behaviour that you're talking about. You can condition them with two stimuli, and they’ll respond to one in almost the Pavlovian way. So, is that a form of consciousness and of intelligence that exists in those inorganic materials?

[Jose] If I could answer your question I wouldn’t be here, I would be at Harvard! (Laughs) The question is, if conscience is the product of biological evolution, this situation would be the product of the biologically evolved conscience producing another system which might develop conscience, but it would be a different sort of conscience.

At the end of the day, I'm wondering whether there are many film references here. I'm thinking about movies in which the machines develop their own conscience, 2001 is the obvious one.

[Audience Member] You made mention of the potential for pain management in future research. Would it be right to say that conditioning and then associative learning are similar to placebo medicine? And, if that is the case, why do we have some placebo medicines the produce feeling even if I haven’t been conditioned?

[Jose] Well, this is a difficult question. Conditioning is part of the mechanisms that produce the placebo effect and I've been I've been addressing that in some lectures but I don't know much about it and I haven't thought about it. But clearly there is an effect of learning and treatment with drugs and signals. The signals maintain the properties of an effective, conditioned stimulus, and then the signal can replace the drug. How do we articulate that and how do we put together these with your interests and mine? I don't know but we can have a chat.

[Paul] Thank you for the questions. There's lots of questions as always so we have a reception afterwards that will give you the opportunity to ask Jose more. Can I now ask Denise to give the formal vote of thanks?

[Denise] Thanks Paul, and a massive thanks to you Jose for an incredibly interesting and stimulating (if I may use that term) talk this evening. You've answered questions that I didn't know I'd got, and you have presented it in a way which people have found able to understand and it's been really interesting and engaging.

I will be (in the good tradition of income generation) selling tickets to the laboratory if anybody would like to join me on a guided tour! There are a number of questions which I still have remaining, not least of all, how the heck do you get in a honeybee into a harness?

So, Jose, it sounds as though we could have a complete menagerie in the North Tower by the time you finished, but thank you very much, and I think you've started us on the next part of our journey in learning more about your research and the opportunities that lie ahead. Thank you.

The presentation slide and the live video in the bottom right corner fades out and a blue screen is presented with the University of Derby’s 3 hills logo in the middle of the screen, in white.

Inaugural Lecture Series: Professor Jose Prados video

Back to The evolution of learning: towards a phylogenetic epistemology