The University of Southampton
Warning! Your browser is out-of-date and not compatible with this website. Please download a new secure and faster browser to view this website correctly.
Telephone:
+442380592690
Telephone:
R.A.Watson@soton.ac.uk

Dr Richard A. Watson

Academic Staff

Personal homepage
Blog feed

Publications in Google Scholar

News: 

How Evolution can Learn

Watson, R.A. & Szathmary, E (2016) "How Can Evolution Learn" Trends in Ecology and Evolution, 31 (2016), pp. 147-157 (DOI: 10.1016/j.tree.2015.11.009)

New Scientist Magazine:  Intelligent Evolution

Feature article Nature’s brain: A radical new view of evolution (pdf)

Is Evolution more Intelligent than we thought?,

The Conversation: Intelligent design without a creator? Why evolution may be smarter than we thought

Extended Evolutionary Synthesis

'Putting the Extended Evolutionary Synthesis to the Test' project details

New project launch featured in: Science 

Press: Southampton set to update Darwin as part of £7.7m project to expand our understanding of evolution 

Watch this space: I will be hiring for two postdocs on this project (starting ~June 2017) looking at the interaction of developmental and ecological networks on evolutionary processes.

Richard A. Watson is an associate professor in the Agents, Interaction and Complexity research group at the University of Southampton's School of Electronics and Computer Science, and a member of the Institute for Life Science, Southampton. He received his BA in Artificial Intelligence from the University of Sussex in 1990 and then worked in industry for five years. Returning to academia, he chose Sussex again for an MSc in Evolutionary and Adaptive Systems, where he was introduced to evolutionary modeling. His PhD in computer science at Brandeis University (2002) resulted in 22 publications and a dissertation addressing the algorithmic concepts underlying the major transitions in evolution. A postdoctoral position at Harvard University's Department of Organismic and Evolutionary Biology provided training to complement his computer science background. He now has over 100 publications on topics spanning evolutionary biology, evolutionary computation, population genetics, neural networks and computational biology. He is the author of Compositional Evolution: The Impact of Sex, Symbiosis, and Modularity on the Gradualist Framework of Evolution (MIT Press, 2006). 

Research

Research interests

 

Watson, R.A. & Szathmary, E (2016) "How Can Evolution Learn" Trends in Ecology and Evolution, 31 (2016), pp. 147-157 (DOI: 10.1016/j.tree.2015.11.009)

Full publications list


People

Current PhD students

Adam Jackson

Adam Jackson

Social evolution; game theory;"meta games"

defended successfully!

Paul Ryan

Paul Ryan

Social evolution; social niche construction; evolution of individuality

defended: passed no corrections!

Miguel Gonzalez Canudas

Miguel Gonzalez Canudas

Cultural evolution; evolution of horizontal information transfer

Simon Tudge

Simon Tudge

(primary supervisor, Markus Brede)

Social evolution; division of labour games; phenotypic plasticity

submitted!

Daniel Power

Dan Power

Ecological evolution; evolution of community structure

William Hurndall

Billy Hurndall

Origin of evolution (chemical systems); evolution of heritability

Louis Kounios

Louis Kounios

Evolution of evolvability; gene regulation networks; optimisation

Kostas Kouvaris

Kostas Kouvaris

Evolution of evolvability; gene regulation networks; evolution of developmental plasticity

 

Ex-PhD students
Rob Mills

Dr Rob Mills  (now in Lisbon)

genetic algorithms; symbiotic algorithms; "multi-scale search"

Simon Powers

Dr Simon Powers  (now in Lausanne)

Social evolution; "social niche construction"

Adam Davies

Dr Adam Davies

Evolution of structure and function

Chris Cox

Chris Cox  (defended April 2015)

Evolutionary optimisation; "schema grammar"


The Algorithm of Evolution

Our work seeks to answer fundamental questions about the algorithms of biological adaptation, especially evolution. This type of work involves modelling biological processes with computers (computational biology), and has spin-offs for new computational methods (e.g. biologically inspired computation, genetic algorithms, neural networks, distributed optimisation algorithms, complex systems engineering). But the focus is on understanding the algorithmic principles of biological processes rather than on simulating biological data or developing computational applications.

Algorithmic abstraction has a precedence in evolutionary biology that is sometimes forgotten. Specifically, Darwin's theory of evolution by natural selection described an algorithm; random variation and selective retention. His (important) conclusions therefore hold in a way that are independent of the mechanistic details of DNA replication or Mendelian genetics, for example. Since then a lot of work in evolutionary biology adds mechanistic details - but very little work asks whether these mechanisms change the underlying algorithm.

The problem of Evo-devo, evo-eco, and evo-ego

Several hot-topics in contemporary evolutionary biology involve struggles to accommodate a number of crucial features into the conventional model. For example, the interaction of evolution and development (evo-devo), the interaction of evolution and ecology (evo-eco), and the interaction of evolution and individuality as in the major transitions in evolution (which we term, "evo-ego"). These areas are problematic because these developmental, ecological and reproductive feedbacks alter the processes of phenotypic variation, selection and inheritance (respectively) on which evolution depends. The conventional model takes variation, selection and inheritance as axiomatic mechanisms that are fixed through evolutionary time, not as variables. This is why topics such as the evolution of evolvability, niche construction and evolutionary transitions in individuality make little sense in the conventional model.

Modelling how evolution changes itself  -  'internalising' parameters 

We therefore seek to take various assumptions that the conventional model defines as exogenous parameters and provide a new algorithmic framework that explains them as endogenous variables. In one point of view, this is sometimes as simple as adding a 'modifier allele' (e.g. that controls the reproductive mode, inheritance mechanism, the genetic relatedness of interactors etc.) or many modifier alleles (e.g. a gene-regulation matrix that controls phenotypic correlations in development, or a matrix of pay-offs in a social game). Then natural selection acts on these modifier alleles, and these modifier alleles change the way that subsequent evolution by natural selection acts. But thinking about it this way often doesnt shed any light on what's possible and what isnt.

A computer science perspective recognises such structures as 'internal states' or a memory that makes evolutionary trajectories 'path dependent'. For example, the path of evolutionary trajectories through phenotype space is subject to developmental constraints and biases (e.g. gene-regulatory organisation, body plans) that are themselves a product of past selection. Moreover, evolutionary trajectories are also subject to selective conditions from the environment (e.g. environmental niches, ecological relationships) that are themselves a product of past selection. Thus both the variation and the selection part of Darwin's algorithm are not fixed but variables modified by the products of the evolutionary process. It is not obvious that continuing to describe these processes within a conventional model is the most effective way of understanding them.

One working hypothesis we are pursuing is that learning algorithms - processes that improve over time by accumulating knowledge from past experience - are a better fit for this purpose than the conventional model. This, and the more general notion of endogenising evolutionary parameters, underpins many of our activities.


 

Research Areas

Evolutionary developmental biology and the evolution of evolvability

Developmental organisation is a product of evolution, and also defines the variability of phenotypes on which evolution depends. We use learning theory to characterise how developmental organisation, and hence phenotypic correlations, change as a function of past selection. This work shows that natural selection can form a distributed 'developmental memory' of multiple past phenotypes, and recall those phenotypes spontaneously. Louis Kounios is presently working to show that, because such a memory is capable of generalisation, it can facilitate evolvability in novel selective environments, and Kostas Kouvaris utilises principles of learning theory (conditions that alleviate overfitting) to facilitate better evolvability/to counteract canalisation and retain developmental plasticity.

Evolutionary ecology and ecological organisation.

The selective pressures experienced by members of a species are a product of their ecological setting or niche, and this niche is itself a product of their evolution (niche construction). Whereas most studies of niche construction focus on the interaction between a species and its abiotic environmental variables, we study the evolution of ecological communities - the evolution of traits that modify the network of fitness interactions between species. Unlike the network of interactions between genes, ecological networks are not selected for the purpose of increasing ecosystem fitness (because ecosystems are not, in most cases, evolutionary units). However, Dan Power finds that we can use theory of unsupervised correlation learning (learning mechanisms that operate without a global reward function) to characterise changes in ecological organisation. This shows the same kind of distributed memory that we see in the evolution of developmental networks - but in this case these are memories of past ecological configurations, good or bad, rather than memories of good phenotypes ie. that were selected in the past.

Evolution of cooperation, social niche construction, meta-games

Social evolution theory tells us that cooperation can evolve if the benefits of cooperation fall differentially on individuals who pass on the cooperative behaviour. This is quantified as direct benefits or indirect benefits via relatedness of interactors (inclusive fitness theory, kin selection). Relatedness is usually taken to be an exogenous parameter of social evolution, but actually organisms have many traits that can change population structure or otherwise affect whether the individuals they interact with are related or not. Simon Powers developed the notion of 'social niche construction' to capture this idea (endogenising parameters that modify social setting) and modelled the evolution of traits that affect group size (and hence, indirectly, relatedness) as an example. Adam Jackson is formalising these notions under a framework of "meta games" - where individuals have traits that change the game they are playing.  

Major evolutionary transitions, evolution of individuality.

Changing the equilibrium level of cooperation via social niche construction (above) is neat. But what we really want to do is understand how new levels of individuality arise. Paul Ryan is examining how individual traits can change the evolutionary unit - i.e. the level of biological organisation at which fitness differences are heritable. This involves social niche construction of individual traits that create a population bottleneck, for example (one of Godfrey-Smiths dimensions of Darwinian individuality). In a group-fissioning process (rather than an aggregation and dispersal process) this supports heritable fitness differences at the group level. But, its still a bit dull when the result of this is merely that all the groups are homogeneous i.e. full of 'cooperators'. Evolutionary transitions are not just about homoeneous cooperation, they are also about providing complementary functions. Simon Tudge (with Markus Brede) is formalising this by characterising "division of labour games". Here fitness is not maximised by positive assortment (as it is in conventional social dilemmas), because you need multiple diverse phenotypes. There are two solutions that we find in biology - 1) Egalitarian transitions (eg. eukaryote organelles) enable selection on groups containing diverse types (despite the fact that the components wont have exactly the same individual fitness) by implementing some form of policing mechanism that synchronises reproduction, 2) Fraternal transitions (eg. evolution of multicellularity) utilise individuals that have context-sensitive phenotypic plasticity - e.g. so they adopt the phenotype that is the complement of the phenotypes they interact with (despite having homogeneous genotypes). The cool thing is that, when the individuality of the group transitions, the unsupervised correlation learning that characterised the evolution of inter-particle relationships before a transition (see ecological organisation above) is converted into supervised correlation learning after the transition (see developmental organisation above). These relationships between particles become developmental relationships (ie within a unit of selection) and control the 'phenotype' of the group. 

More

  • Evolution of template replication in the origin of life (Billy Hurndall).
  • Evolution of grammars that learn to exploit problem structure (Chris Cox).
  • Evolutionary optimisation inspired by evolutionary transitions (Rob Mills)
  • Evolution of culture (Miguel Gonzalez).
  • Self-modelling dynamical systems - optimisation based on learning not natural selection.
  • Epigenetic inheritance and Lamarckian evolution.
  • Adaptation without natural selection (or design).

Full publications list


 

Research questions

  • Does biological evolution optimise something?

In evolutionary computation, the idea that someone would doubt this is perplexing - we use genetic algorithms to optimise functions routinely (of course, we also appreciate they are local optimisers not global). But in evolutionary theory, its common to describe evolution as a dynamical system - a dynamical system that does not necessarily optimise fitness. In fact, sometimes it minimises it (under freq dependent fitness). In fact, some say that the conditions under which fitness is optimised are a special case. Worse, there is not always any potential function that describes the behaviour of this dynamical system (so there's no obvious quantity that monotonically increases/decreases as a function of ENS). Some claim to be happy with this - 'there is no reason to suppose that evolution optimises anything, its just a process doing whatever it does'. But if so, how is it an explanation for adaptation? Grafen (I think, rightly) equates evolution with an optimisation process not so much because he can show that it is (which of course it is, under some circumstances) - but more because, if you dont do that, its useless for explaining the 'goodness of fit' between organisms and environments that evolution by natural selection was supposed to explain. The general question remains, however: How do we link the description of evolution as a dynamical system with evolution as an explanation for adaptation. 

  • How does evolution by natural selection produce adaptive complexity?

Darwinian evolution does not make any reference to the complexity of entities - only to their ability to survive and reproduce (and simple things can do that). Does evolution produce complexity in order to maximise fitness, or is complexity a non-systematic by-product? What would a theory of complexity look like? (I think the product of evolution has to be a dynamical thing (like a developmental process) not a static thing (like a phenotype). And it has to have multiple-levels of organisation, with localised autonomy... maybe). 

  • Can evolution increase its ability to adapt over time (evolution of evolvabiity)?

Some lineages may have properties (eg genetic or developmental organisation, body plans, etc.) that enable better adaptation (reproduction or survival) in the long term than others. But evolution is myopic - it favours things that are fit now; not things that could be fit sometime in the future. So these long-term properties of lineages cannot be adaptations - they can only be happy accidents. (Or properties that survive 'after the fact'). But... what is the relationship between adaptations that confer fitness advantages in the short-term and these long-term adaptive consequences? Note that learning mechanisms fit a model to training data (in the past) but test the model on novel data (in the future). Successful generalisation doesnt need the future to cause the past, however - it is therefore not impossible. This suggests that the evolution of long-term evolvability from short-term selection is not impossible either. But successful generalisation only works when there are structural regularities that are invariant across both sets, and the model class is capable of representing those structural regularities, and conditions are such that the model doesnt overfit. This gives us a formal framework for characterising conditions for the evolution of evolvability from short term selection.

  • What principles govern the evolution of biological structure rather than function (organisation rather than fitness)?

Darwinian evolution is a theory about fitness (survival and reproduction); it makes no reference to organisation (except that fit organisations are favoured over unfit ones). What would a theory of biological organisation look like? How do organisation and fitness interact? e.g. Does the evolution of modularity in gene-regulation networks reflect modularity in the selective environment? Is it a 'model' of the 'problem space'?

  • Can ecosytems be 'fit'? Or non-arbitrarily organised in any systematic way?

Since an ecosystem is not an evolutionary unit, is the organisation of an ecosystem necessarily arbitrary (i.e. disorganised)? Does the evolution of community structure in ecological networks reflect anything about the ecological environment/conditions? Can unsupervised learning theory (learning without a reward function) help us understand the evolution of biological organisations that are not evolutionary units?

  • How do structures that facilitate robustness (canalisation) relate to structures that facilitate flexibility (evolvability)?

Are they opposed to one another or are they synergistic? Not producing bad things might increase the likelihood of producing good things. But the easiest way to not produce bad things, is to not vary at all. Note that in learning systems, the parameters of a model are adjusted to reduce error on the training set - if the model underfits error will remain high. If the model fits well - generalisation will be good. But if the model overfits, it  will reduce error on the training set but fail to generalise well on the test. The evolutionary analogue of this is canalisation that improves the average fitness of offspring inthe short term, but prevents future flexibility and the possibility of adaptation. The conditions that alleviate overfitting in learning systems inform us of conditions that resolve the tension between canalisation and evolvability.

  • What kind of problem can evolution solve?

Is it only the same kind of problems that a hill-climber can solve? (including noisy/stochastic/restart hill climber). Can evolution (implicitly) do things like problem decomposition, hierarchical problem abstraction, representational recoding? Is that what the evolution of develomental modularity is doing? Can evolution implement short-term 'goal-setting' (e.g. changing the effective selection pressure experienced through niche construction) that facilitates long-term Darwinian fitness?

  • Is there an algorithm of biological evolution that we have not yet identified?

One that contains evolution by natural selection as Darwin described it, but isnt ENS? (like matrix multiplication contains addition but isnt addition)? One that incorporates the evolution of developmental, ecological and reproductive organisation? One that connects with the issues articulated by the 'Extended Evolutionary Synthesis'? (developmental plasticity, niche construction, epigenetic inheritance). I'm working on a model of 'deep evolution' based on 'deep optimisation' (aka multi-scale search) inspired by 'deep learning' in neural networks.

  • How do the major transitions in evolution work?

They involve new reproductive organisations (vertical transmission, genetic linkage, compartmentalisation) that create/support heritable fitness differences at a new level of organisation. But how can the adaptations necessary for this new level of organisation evolve before that level of organisation exists as an evolutionary unit? They must be adaptations that increase the fitness of the existing evolutionary units. But if so, why would they ever create new units that oppose their own interests? And if the new structures dont oppose the selective interests of the existing units - how can they alter evolutionary outcomes/have any evolutionary significance? Ans: see the use of unsupervised learning to guide deep learning - i.e. unsupervised learning is useful for 'priming' good performance at the next layer of the network (even though the next layer of the network doesnt exist yet) because reducing the dimensionality of the data is generally a good idea regardless of what you're going to use it for.

  • How does social behaviour and cooperation evolve?

In particular, we are interested in coordinated cooperative behaviours that create adaptive complexity at new levels of biological organisation. How do individuals end-up playing the social game that they are playing? Can they change it? ("meta games"). Can individuals change the structure of social interactions through natural selection? The evolution of population structure (including relatedness) ("social niche construction"). How does division of labour evolve? (which is tricky because maintaining heterogeneous roles (that are unlikely to be exactly the same fitness under individual/particle-level selection) is problematic/unstable.

Research history

I've worked on genetic algorithms, coevolutionary algorithms, the benefit of sexual recombination, the major evolutionary transitions, fitness landscapes, artificial life models, dynamical systems, modularity, model-building optimisation, modularity, multi-objective optimisation, optimisation, pareto coevolution, population genetics, collective robotics....

See full publications

News and Announcements

  • How Evolution can Learn

Watson, R.A. & Szathmary, E (2016) "How Can Evolution Learn" Trends in Ecology and Evolution, 31 (2016), pp. 147-157 (DOI: 10.1016/j.tree.2015.11.009)

New Scientist Magazine: https://d1o50x50snmhul.cloudfront.net/wp-content/uploads/2016/03/nsc_20160326-800x1052.jpg Intelligent Evolution

Feature article Nature’s brain: A radical new view of evolution (pdf)

Is Evolution more Intelligent than we thought?,

The Conversation: Intelligent design without a creator? Why evolution may be smarter than we thought

  • Extended Evolutionary Synthesis

'Putting the Extended Evolutionary Synthesis to the Test' project details

New project launch featured in: Science 

Press: Southampton set to update Darwin as part of £7.7m project to expand our understanding of evolution 

Watch this space: I will be hiring for two postdocs on this project (starting ~June 2017) looking at the interaction of developmental and ecological networks on evolutionary processes."To Group or not to Group" is a perspective in Science (Dec 2011) by Eors Szathmary discussing the paper "The Concurrent Evolution of Cooperation and the Population Structures that Support it."

Grants

'Putting the Extended Evolutionary Synthesis to the Test' (£7.7M) - started Oct 2016

Spring School in Complexity Science, 2006. EPSRC.

CTA scholarships for MSc in Complexity Science.

"Principles of Distributed Intelligence in Self-Organised Multi-Agent Networks" - DSTLX-1000074615

Professional

Qualifications

PhD "Compositional Evolution" Brandeis University 2002
MSc "Knowledge Based Systems" Sussex University 1996
BA "Computing with Artificial Intelligence" Sussex University 1990

Duties

Programme leader MSc Artificial Intelligence

Publications

Watson, Richard A. and Pollack, Jordan B. (2002) A Computational Model of Symbiotic Composition in Evolutionary Transitions. Biosystems, 69, (2-3), 187-209.

Watson, Richard A. (2006) Compositional Evolution: The impact of Sex, Symbiosis and Modularity on the Gradualist Framework of Evolution, MIT Press (Vienna series in theoretical biology, NA).

Pollack, Jordan B., Bedau, Mark, Husbands, Phil, Ikegami, Takashi and Watson, Richard A., Pollack, Jordan B., Bedau, Mark, Husbands, Phil, Ikegami, Takashi and Watson, Richard A. (eds.) (2004) Artificial Life IX: Proceedings of the Ninth International Conference on the Simulation and Synthesis of Living Systems , MIT Press

Watson, Richard A., Ficici, Sevan G. and Pollack, Jordan B. (2002) Embodied Evolution: Distributing an evolutionary algorithm in a population of robots. Robotics and Autonomous Systems, 39, (1), 1-18.

Watson, Richard A. and Pollack, Jordan B. (2005) Modular Interdependency in Complex Dynamical Systems. Artificial Life, 11, (4), 445-457.

Silverman, Eric and Bullock, Seth (2004) Empiricism in artificial life. In, the Ninth International Conference on Artificial Life MIT Press, Cambridge, MA, 534-539.

Mills, Rob and Watson, Richard A., Capcarrere, Mathieu S., Freitas, Alex A., Bentley, Peter J., Johnson, Colin G. and Timmis, Jon (eds.) (2005) Genetic Assimilation and Canalisation in the Baldwin Effect. Proceedings of 8th European Conference on Artificial Life (ECAL 2005), LNCS 3, 353-362.

Lenaerts, Tom, Chu, Dominique and Watson, Richard A. (2005) Dynamical Hierarchies. Artificial Life, 11, (4), 403-405.

Watson, Richard A. and Wakeley, John (2005) Multidimensional Epistasis and the Advantage of Sex. Proceedings of the 2005 Congress on Evolutionary Computation (CEC 2005), 3, 2792-2799.

Weinreich, Daniel M., Watson, Richard A. and Chao, Lin (2005) Perspective:Sign Epistasis and Genetic Constraint on Evolutionary Trajectories. Evolution, 59, (6), 1165-1174.

Watson, Richard A. (2005) On the Unit of Selection in Sexual Populations. In Advances in Artificial Life, Eighth European Conference (ECAL 2005)

Mills, Rob and Watson, R. A., Rocha, Luis M., Bedau, Mark, Floreano, Dario, Goldstone, Robert, Vespignani, Alessandro and Yaeger, Larry (eds.) (2006) On Crossing Fitness Valleys with the Baldwin Effect. Proceedings of the Tenth International Conference on the Simulation and Synthesis of Living Systems, 493-499.

Watson, Richard A., Yao, Xin and al, et (eds.) (2004) A Simple Two-Module Problem to Exemplify Building-Block Assembly Under Crossover. Proceedings of the 8th International Conference on Parallel Problem Solving from Nature (PPSN-VIII), 161-171.

Watson, Richard A. (2002) Compositional Evolution: Interdisciplinary Investigations in Evolvability, Modularity, and Symbiosis. Brandeis University, MA. USA, Computer Science, Doctoral Thesis .

Watson, Richard A. (2001) Analysis of Recombinative Algorithms on a Non-Separable Building-Block Problem. In, Martin, Worthy N. and Spears, William M. (eds.) Foundations of Genetic Algorithms, Volume 6. , Morgan Kaufmann, 69-90.

Watson, Richard A. and Pollack, Jordan B., Kelemen, Jozef and Sos, Petr (eds.) (2001) Symbiotic Composition and Evolvability. Proceedings of European Conference on Artificial Life (ECAL 2001), 480-490.

Watson, Richard A. and Pollack, Jordan B., Floreano, Dario, Nicoud, Jean-Daniel and Mondada, Francesco (eds.) (1999) How Symbiosis Can Guide Evolution. Proceedings of the 5th European Conference on Advances in Artificial Life (ECAL 1999), 29-38.

Watson, Richard A. and Pollack, Jordan B., Spector, Lee (eds.) (2001) Coevolutionary Dynamics in a Minimal Substrate. Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2001), 702-709.

Watson, Richard A., Reil, Torsten and Pollack, Jordan B., Bedau, M., McCaskill, J., Packard, N. and Rasmussen, S. (eds.) (2000) Mutualism, Parasitism, and Evolutionary Adaptation. Proceedings of Artificial Life VII (ALife VII), 170-178.

Watson, Richard A., Hornby, Gregory S. and Pollack, Jordan B., Eiben, A. E., Back, T., Schoenauer, M. and Schweffel, H.-B. (eds.) (1998) Modeling Building Block Interdependency. Proceedings of Parallel Problem Solving from Nature V (PPSN V), 97-106.

Dauscher, Peter, Polani, Daniel and Watson, Richard A., Rocha, M., Bedau, M., Floreano, D., Goldstone, R., Vespignani, A. and Yaeger, L. (eds.) (2006) A Simple Modularity Measure for Search Spaces based on Information Theory. Proceedings of Artificial Life X (ALife X)

De Jong, Edwin, Watson, Richard A. and Thierens, Dirk (2005) On the Complexity of Hierarchical Problem Solving. Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2005)

Watson, Richard A., Weinreich, Daniel and Wakeley, John (2006) Effects of intra-gene fitness interactions on the benefit of sexual recombination. Biochemical Society Transactions, 34, (4), 560-561.

Mills, Rob and Watson, Richard A., Thierens, Dirk and Lipson, Hod (eds.) (2007) Variable Discrimination of Crossover Versus Mutation Using Parameterized Modular Structure. Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2007), 1312-1319.

Mills, Rob and Watson, Richard A., Almeida e Costa, Fernando, Rocha, Luis M., Costa, Ernesto, Harvey, Inman and Coutinho, Antonio (eds.) (2007) Symbiosis, Synergy and Modularity: Introducing the Reciprocal Synergy Symbiosis Algorithm. Proceedings of 9th European Conference on Artificial Life (ECAL 2007), 1192-1201.

Powers, Simon T. and Watson, Richard A. (2007) Preliminary Investigations into the Evolution of Cooperative Strategies in a Minimally Spatial Model. At Proceedings of the 9th annual conference on Genetic and evolutionary computation (GECCO 2007), London, 07 - 11 Jul 2007. ACM Press, 343-343.

Powers, Simon T. and Watson, Richard A. (2007) Investigating the Evolution of Cooperative Behaviour in a Minimally Spatial Model. In, e Costa, Fernando A., Rocha, Luis M., Costa, Ernesto, Harvey, Inman and Coutinho, António (eds.) Advances in Artificial Life : Proceedings of the Ninth European Conference on Artificial Life (ECAL 2007). the 9th European Conference on Artificial Life (ECAL 2007) , Springer , 605-614.

Powers, Simon T., Penn, Alexandra S. and Watson, Richard A. (2007) Individual Selection for Cooperative Group Formation. In, e Costa, Fernando A., Rocha, Luis M., Costa, Ernesto, Harvey, Inman and Coutinho, António (eds.) Advances in Artificial Life: Proceedings of the Ninth European Conference on Artificial Life (ECAL 2007). the 9th European Conference on Artificial Life (ECAL 2007) , Springer , 585-594.

Watson, Richard A. and Jansen, Thomas, Thierens, Dirk and Lipson, Hod (eds.) (2007) A Building-Block Royal Road Where Crossover is Provably Essential. Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2007), 1452-1459.

Noble, J. and Watson, R. A. (2001) Pareto coevolution: Using performance against coevolved opponents in a game as dimensions for Pareto selection. In, Spector, L., Goodman, E., Wu, A., Langdon, W. B., Voigt, H.-M., Gen, M., Sen, S., Dorigo, M., Pezeshk, S., Garzon, M. and Burke, E. (eds.) Proceedings of the Genetic and Evolutionary Computation Conference, GECCO-2001. , Morgan Kauffman, 493-500.

Powers, Simon T. and Watson, Richard A. (2008) The Group Selection Debate and ALife: Weak Altruism, Strong Altruism, and Inclusive Fitness (abstract). In, Artificial Life XI: Proceedings of the Eleventh International Conference on the Simulation and Synthesis of Living Systems. The Eleventh International Conference on the Simulation and Synthesis of Living Systems (Alife XI) , MIT Press, Cambridge, MA, 796.

Mills, Rob and Watson, Richard A. (2008) Adaptive units of selection can evolve complexes that are provably unevolvable under fixed units of selection (abstract). In, Bullock, S., Noble, J. , Watson, R. A. and Bedau, M. A. (eds.) Artificial Life XI: Proceedings of the Eleventh International Conference on the Simulation and Synthesis of Living Systems. 11th International Conference on the Simulation and Synthesis of Living Systems , MIT Press, 785.

Geard, Nicholas and Bullock, Seth (2008) Group formation and social evolution: a computational model. In, Bullock, S., Noble, J., Watson, R. and Bedau, M.A. (eds.) Artificial Life XI: Proceedings of the Eleventh International Conference on the Simulation and Synthesis of Living Systems. The Eleventh International Conference on the Simulation and Synthesis of Living Systems (Artificial Life XI) Cambridge, GB, MIT Press, 197-203.

Jacyno, Mariusz and Bullock, Seth (2008) Energy, entropy and work in computational ecosystems: a thermodynamic account. In, Bullock, S., Noble, J., Watson, R. and Bedau, Mark A. (eds.) Artificial Life XI: Proceedings of the Eleventh International Conference on the Simulation and Synthesis of Living Systems. Eleventh International Conference on Artificial Life Cambridge, GB, MIT Press, 274-281.

Powers, Simon T, Penn, Alexandra S and Watson, Richard A (2008) The Efficacy of Group Selection is Increased by Coexistence Dynamics within Groups. In, Bullock, Seth , Noble, Jason and Watson, Richard (eds.) Artificial Life XI: Proceedings of the Eleventh International Conference on the Simulation and Synthesis of Living Systems. The Eleventh International Conference on the Simulation and Synthesis of Living Systems , MIT Press, Cambridge, MA, 498-505.

Groom, Graeme, Mills, Rob and Watson, Richard A. (2008) How epigenetic evolution can guide genetic evolution (abstract). In, Bullock, Seth, Noble, Jason, Watson, Richard A. and Bedau, M. A. (eds.) Artificial Life XI: Proceedings of the Eleventh International Conference on the Simulation and Synthesis of Living Systems. Proceedings of the Eleventh International Conference on the Simulation and Synthesis of Living Systems , MIT Press, Cambridge, MA, 772.

Watson, Richard A., Mills, Robert, Penn, Alexandra and Powers, Simon T (2008) Can individual selection favour significant higher-level selection? (abstract). At The Eleventh International Conference on the Simulation and Synthesis of Living Systems (Alife XI), Winchester, UK, 05 - 08 Aug 2008. MIT Press, 818-818.

James, E., Noble, J. and Watson, R. (2008) Solving the division of labour problem using stigmergy and evolved heterogeneity (abstract). In, Bullock, S., Noble, J., Watson, R. and Bedau, M. A. (eds.) Artificial Life XI: Proceedings of the Eleventh International Conference on the Simulation and Synthesis of Living Systems. , MIT Press, Cambridge, MA, 778.

Bullock, S., Noble, J., Watson, R. and Bedau, M. (2008) Artificial Life XI: Proceedings of the Eleventh International Conference on the Simulation and Synthesis of Living Systems, MIT Press

Penn, Alexandra S, Powers, Simon T, Conibear, Tim , Kraaijeveld, Alex , Watson, Richard , Bigg , Zoe and Webb, Jeremy (2008) Co-operation and Group structure in Bacterial Biofilms. At Society for General Microbiology, Autumn meeting. , Trinity College, Dublin,

Watson, Richard A., Buckley, C. L. and Mills, Rob (2009) The Effect of Hebbian Learning on Optimisation in Hopfield Networks.

Mills, Rob and Watson, Richard A., Kampis, George and Szathmáry, Erös (eds.) (2009) Symbiosis Enables the Evolution of Rare Complexes in Structured Environments. Proceedings of 10th European Conference on Artificial Life (ECAL 2009), 110-117.

Powers, Simon T and Watson, Richard A (2009) Evolution of Individual Group Size Preference can Increase Group-level Selection and Cooperation. In, Advances in Artificial Life: Proceedings of the Tenth European Conference on Artificial Life (ECAL 2009). , Springer.

Watson, Richard A., Buckley, C. L. and Mills, Rob (2009) Global Adaptation in Networks of Selfish Components: Emergent Associative Memory at the System Scale.

Watson, Richard, Palmius, Niclas, Mills, Robert, Powers, Simon T and Penn, Alexandra, Kampis, George and Szathmáry, Erös (eds.) (2009) Can Selfish Symbioses Effect Higher-level Selection? Proceedings of 10th European Conference on Artificial Life (ECAL 2009)

Snowdon, James, Powers, Simon T and Watson, Richard (2009) Moderate contact between sub-populations promotes evolved assortativity enabling group selection. At Proceedings of 10th European Conference on Artificial Life (ECAL 2009) Springer.

Powers, Simon T, Mills, Rob, Penn, Alexandra S and Watson, Richard A (2009) Social Environment Construction Provides an Adaptive Explanation for New Levels of Individuality. At Levels of Selection and Individuality in Evolution: Conceptual Issues and the Role of Artificial Life Models, Budapest, Hungary, , 18-21.

Watson, Richard A., Buckley, C. L. and Mills, Rob (2011) Optimisation in ‘Self-modelling’ Complex Adaptive Systems. Complexity, 16, (5), 17-26.

Watson, Richard, Weinreich, Daniel M. and Wakeley, John (2011) Genome structure and the benefit of sex. Evolution, 65, (2), 523-536. (doi:10.1111/j.1558-5646.2010.01144.x). (PMID:21029076).

Watson, Richard A., Mills, Rob and Buckley, C.L. (2011) Global adaptation in networks of selfish components: emergent associative memory at the system scale. Artificial Life, 17, (3), Summer Issue, 147-166. (doi:10.1162/artl_a_00029).

Powers, Simon T., Penn, Alexandra S. and Watson, Richard A. (2011) The Concurrent Evolution of Cooperation and the Population Structures that Support it. Evolution, 65, (6), 1527-1543. (doi:10.1111/j.1558-5646.2011.01250.x).

Watson, Richard, Mills, Rob, Buckley, C. L., Penn, Alexandra, Davies, Adam, Noble, Jason and Bullock, Seth (2010) Adaptation without natural selection. In, Fellerman, Harold, Dörr, Mark, Hanczyc, Martin M., Ladegaard Laursen, Lone, Maurer, Sarah, Merkle, Daniel, Monnard, Pierre-Alain, Stoy, Kasper and Rasmussen, Steen (eds.) Artificial Life XII: Proceedings of the Twelfth International Conference on the Synthesis and Simulation of Living Systems. , MIT Press, 80-81.

Noble, Jason, Hebbron, Tom, Van Der Horst, Johannes, Mills, Rob, Powers, Simon T. and Watson, Richard (2010) Selection pressures for a theory-of-mind faculty in artificial agents. In, Fellerman, Harold, Dörr, Mark, Hanczyc, Martin M., Ladegaard Laursen, Lone, Maurer, Sarah, Merkle, Daniel, Monnard, Pierre-Alain, Stoy, Kasper and Rasmussen, Steen (eds.) Artificial Life XII: Proceedings of the Twelfth International Conference on the Synthesis and Simulation of Living Systems. , MIT Press, 615-615.

Davies, Adam, Watson, Richard, Mills, Rob, Buckley, C. L. and Noble, Jason (2010) If you can't be with the one you love, love the one you're with: How individual habituation of agent interactions improves global utility. In, Fellerman, Harold, Dörr, Mark, Hanczyc, Martin M., Ladegaard Laursen, Lone, Maurer, Sarah, Merkle, Daniel, Monnard, Pierre-Alain, Stoy, Kasper and Rasmussen, Steen (eds.) Artificial Life XII: Proceedings of the Twelfth International Conference on the Synthesis and Simulation of Living Systems. , MIT Press, 659-666.

Watson, Richard A., Mills, Rob and Buckley, C. L. (2011) Transformations in the Scale of Behaviour and the Global Optimisation of Constraints in Adaptive Networks. Adaptive Behavior, 19, (4), 227-249.

Mills, Rob, Watson, Richard A. and Buckley, Christopher L. (2011) Emergent associative memory as a local organising principle for global adaptation in adaptive networks. At Eighth International Conference on Complex Systems, Boston, MA, , 417-430.

Mills, Rob and Watson, Richard A. (2011) Multi-scale search, modular variation, and adaptive neighbourhoods. .

McCabe, Connor, Watson, Richard, Prichard, Jane S. and Hall, Wendy (2011) The Web as an Adaptive Network: Coevolution of Web Behavior and Web Structure. In, Web Science Conference 2011, Koblenz , Germany, 14 - 17 Jun 2011.

Davies, Adam, Watson, Richard, Mills, Rob, Buckley, C.L. and Noble, Jason (2011) "If you can't be with the one you love, love the one you're with": How individual habituation of agent interactions improves global utility. Artificial Life, 17, (3), Summer Issue, 167-181. (doi:10.1162/artl_a_00030).

Powers, Simon T, Heys, Christopher and Watson, Richard A (2011) How to Measure Group Selection in Real-world Populations. In, Advances in Artificial Life, ECAL 2011. , MIT Press, 672-679.

Penn, Alexandra, S., Conibear, Tim C.R., Watson, Richard, A., Kraaijeveld, Alex R. and Webb, Jeremy, S. (2012) Can Simpson's Paradox explain co-operation in Pseudomonas aeruginosa biofilms? [in special issue: Biofilms II] FEMS Immunology & Medical Microbiology, 65, (2), 226-235. (doi:10.1111/j.1574-695X.2012.00970.x).

Weinreich, Daniel M., Sindi, Suzanne and Watson, Richard A. (2012) Finding the boundary between evolutionary basins of attraction, and implications for Wright’s fitness landscape analogy. [in special issue: Statistical Mechanics and the Dynamics of Evolution] Journal of Statistical Mechanics Theory and Experiment, 523-536.

Watson, Richard, Buckley, C. L., Mills, Rob and Davies, Adam (2010) Associative memory in gene regulation networks. In, Fellerman, Harold, Dörr, Mark, Hanczyc, Martin M., Ladegaard Laursen, Lone, Maurer, Sarah, Merkle, Daniel, Monnard, Pierre-Alain, Stoy, Kasper and Rasmussen, Steen (eds.) Artificial Life XII: Proceedings of the Twelfth International Conference on the Synthesis and Simulation of Living Systems. , MIT Press, 659-666. (Proceedings of the Twelth International Conference on the Simulation and Synthesis of Living Systems).

Watson, Richard (2012) Is evolution by natural selection the algorithm of biological evolution? In, Adami, Christoph, Bryson, David M., Ofria, Charles and Pennock, Robert T. (eds.) Artificial Life XIII: Proceedings of the Thirteenth International Conference on the Synthesis and Simulation of Living Systems. Artificial Life XIII: 13th International Conference on the Simulation and Synthesis of Living Systems , MIT Press, 121-128. (doi:10.7551/978-0-262-31050-5-ch018).

Watson, Richard (2012) Individual and global adaptation in networks. In, Wróbel, Borys (eds.) Artificial Life XIII: Proceedings of the Thirteenth International Conference on the Synthesis and Simulation of Living Systems. EVONETS Workshop at ALife XIII: 13th International Conference on Artificial Life , MIT Press.

Penn, Alexandra S, Conibear, Tim C.R., Watson, Richard A., Kraaijeveld, Alex R. and Webb, Jeremy S. (2012) Can Simpson's paradox explain co-operation in Pseudomonas aeruginosa biofilms? FEMS Immunology & Medical Microbiology, 65, (2), 226-235. (doi:10.1111/j.1574-695X.2012.00970.x).

Doncaster, C. Patrick, Jackson, Adam and Watson, Richard A. (2013) Manipulated into giving: when parasitism drives apparent or incidental altruism. Proceedings of The Royal Society B. Biological Sciences, 280, (1758), 20130108. (doi:10.1098/rspb.2013.0108). (PMID:23486440).

Doncaster, C. Patrick, Jackson, Adam and Watson, Richard A. (2013) Competitive environments sustain costly altruism with negligible assortment of interactions. Scientific Reports, 3, (2836), 1-6. (doi:10.1038/srep02836).

Gonzalez, Miguel, Watson, Richard A., Noble, Jason and Bullock, Seth (2014) The origin of culture: Selective conditions for horizontal information transfer. In, Lipson, Hod, Sayama, Hiroki, Rieffel, John, Risi, Sebastian and Doursat, Rene (eds.) ALIFE 14: The Fourteenth International Conference on the Synthesis and Simulation of Living Systems. Fourteenth International Conference on Artificial Life , MIT Press.

Calman, L., Redfern, SJ and Watson, Richard A. (2000) Competence to practice of student nurses: theoretical frameworks, instruments and concerns. In, The Royal College of Nursing Research Society Annual Nursing Research Conference, Sheffield, GB,

Doncaster, C.P., Jackson, A. and Watson, R.A. (2013) Manipulated into giving: when parasitism drives apparent or incidental altruism. Proceedings of The Royal Society B Biological Sciences, 280, (1758), 20130108-[10pp]. (doi:10.1098/rspb.2013.0108). (PMID:23486440).

Watson, Richard, Mills, Rob, Buckley, C.L., Kouvaris, Konstantinos, Jackson, Adam, Powers, Simon T. , Cox, Chris, Tudge, Simon, Davies, Adam, Kounios, Loizos and Power, Daniel (2015) Evolutionary connectionism: algorithmic principles underlying the evolution of biological organisation in dvo-devo, evo-eco and evolutionary transitions. Evolutionary Biology, 1-31. (doi:10.1007/s11692-015-9358-z).

Tudge, Simon, Brede, Markus and Watson, Richard (2013) Cooperation and the division of labour. In, Proceedings of the twelfth European Conference on the Synthesis and Simulation of Living Systems, Taormina, IT, 02 - 06 Sep 2013. 8pp. (doi:10.7551/978-0-262-31709-2-ch001).

Tudge, Simon, Brede, Markus and Watson, Richard (2015) The evolution of assortment with multiple simultaneous games. In, The European Conference on Artificial Life 2015, York, GB, 20 - 24 Jul 2015. 7pp.

Tudge, Simon, Brede, Markus, Watson, Richard and Gonzalez, Miguel (2014) Hamilton’s rule in non-additive games. Author's original, 1-11.

Power, Daniel A., Watson, Richard A., Szathmáry, Eörs, Mills, Rob, Powers, Simon T., Doncaster, C. Patrick and Czapp, Blazej (2015) What can ecosystems learn? Expanding evolutionary ecology with learning theory. Biology Direct, 10, (1), 1-24. (doi:10.1186/s13062-015-0094-1).

Watson, Richard and Ebner, Marc (2013) Eco-evolutionary dynamics on deformable fitness landscapes. In, Richter, Hendrik and Engelbrecht, Andries (eds.) Recent Advances in the Theory and Application of Fitness Landscapes. Berlin, DE, Springer, 339-368. (Emergence, Complexity and Computation, 6). (doi:10.1007/978-3-642-41888-4).

Watson, Richard and Szathmary, Eors (2015) How can evolution learn? Trends in Ecology & Evolution, 31, (2), 147-157. (doi:10.1016/j.tree.2015.11.009).

Watson, Richard A., Wagner, Gunter, Pavlicev, Mihaela, Weinreich, Daniel M. and Mills, Rob (2014) The evolution of phenotypic correlations and “developmental memory”. Evolution, 68, (4), 1124-1138. (doi:10.1111/evo.12337).

Tudge, Simon J., Watson, Richard A. and Brede, Markus (2016) Game theoretic treatments for the differentiation of functional roles in the transition to multicellularity. Journal of Theoretical Biology, 395, 161-173. (doi:10.1016/j.jtbi.2016.01.041).

Contact

Telephone: +442380592690

Email: R.A.Watson@soton.ac.uk

Fax: + 44 (0) 23 8059 2865

Additional contact details

Location: Bld 32, Rm. 4035

Postal address: 
School of Electronics and Computer Science
University of Southampton, Highfield
Southampton, SO17 1BJ - U.K.

Share this profile FacebookGoogle+TwitterWeibo

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×