Tag Archives: emergent behaviour

Blind to complexity

fruit fly nervous system Albert Cardona HHMI Janelia Research Campus Welcome Image Awards 2015When faced with complexity, we tend to seek order and simplicity.  Most of us respond negatively to the uncertainty associated with complex systems and their apparent unpredictability.  Complex systems can be characterised as large networks operating using simple rules but without central control which results in self-organising behaviour and non-trivial emergent behaviour.  Emergent behaviour is the behaviour of the system that is not apparent or expected from the behaviour of its constituent parts [see ‘Emergent properties‘ on September 16th, 2015].

The philosopher, William Wimsatt observed that we tend to ignore phenomena whose complexity exceeds our predictive capability and our detection apparatus.  This is problematic because we try to over-simplify our descriptions of complex systems.  Occam’s razor is often mis-interpreted to mean that simple explanations are better ones, whereas in reality ‘everything should be made as simple as possible, but not simpler’, (which is often attributed to Einstein).  This implies that our explanation and any mathematical model of a complex system, such as the nervous system in the image, will need to be complex.  In mathematical terms, this will probably mean a non-linear dynamic model with a solution in the form of a phase portrait.  ‘Non-linear’ because the response of the system not proportional to the stimulus inducing the response; ‘dynamic’ because the system changes with time; and a ‘phase portrait’ because the system can exist in many states, some stable and some unstable, dependent on its prior history; so, for instance for a pendulum, its phase portrait is a plot of all of its possible positions and velocities.

If all this sounds too hard, then you see why people shy away from using complex models to describe a complex system even when it is obvious that the system is complex and extremely unlikely to be adequately described by a linear model, such as for the nervous system in the image.

In other words, if we can’t see it and its too hard to think about it, then we pretend it’s not happening!

 

The thumbnail shows an image of a fruit-fly’s nervous system taken by Albert Cardona from HHMI Janelia Research Campus.  The image won a Wellcome Image Award in 2015.

William C. Wimsatt, Randomness and perceived randomness in evolutionary biology, Synthese, 43(2):287-329, 1980.

For more on this topic see: ‘Is the world comprehensible?‘ on March 15th, 2017.

 

Red to blue

Some research has a very long incubation time.  Last month, we published a short paper that describes the initial results of research that started just after I arrived in Liverpool in 2011.  There are various reasons for our slow progress, including our caution about the validity of the original idea and the challenges of working across discipline boundaries.  However, we were induced to rush to publication by the realization that others were catching up with us [see blog post and conference paper].  Our title does not give much away: ‘Characterisation of metal fatigue by optical second harmonic generation‘.

Second harmonic generation or frequency doubling occurs when photons interact with a non-linear material and are combined to produce new photons with twice the energy, and hence, twice the frequency and half the wavelength of the original photons.  Photons are discrete packets of energy that, in our case, are supplied in pulses of 2 picoseconds from a laser operating at a wavelength of 800 nanometres (nm).  The photons strike the surface, are reflected, and then collected in a spectrograph to allow us to evaluate the wavelength of the reflected photons.  We look for ones at 400 nm, i.e. a shift from red to blue.

The key finding of our research is that the second harmonic generation from material in the plastic zone ahead of a propagating fatigue crack is different to virgin material that has experienced no plastic deformation.  This is significant because the shape and size of the crack tip plastic zone determines the rate and direction of crack propagation; so, information about the plastic zone can be used to predict the life of a component.  At first sight, this capability appears similar to thermoelastic stress analysis that I have described in Instructive Update on October 4th, 2017; however, the significant potential advantage of second harmonic generation is that the component does not have to be subject to a cyclic load during the measurement, which implies we could study behaviour during a load cycle as well as conduct forensic investigations.  We have some work to do to realise this potential including developing an instrument for routine measurements in an engineering laboratory, rather than an optics lab.

Last week, I promised weekly links to posts on relevant Thermodynamics topics for students following my undergraduate module; so here are three: ‘Emergent properties‘, ‘Problem-solving in Thermodynamics‘, and ‘Running away from tigers‘.

 

Digital hive mind

durham-cloistersFor many people Durham Cathedral will be familiar as a location in the Harry Potter movies.  However, for me it triggers memories of walking around the cloisters discussing Erwin Schrodinger’s arithmetical paradox: there seems to be a great number of conscious egos creating their own worlds but only one world.  Each of us appears to construct our own domain of private consciousness and Schrodinger identifies the region where they all overlap as the ‘real world around us’.  However, he raises questions such as, is my world really the same as yours?  Schrodinger proposes two solutions to the paradox: either there are a multitude of worlds with no communication between them or a unification of minds or consciousness.

Schrodinger found ‘it utterly impossible to form an idea about’ how his ‘own conscious mind should have originated by the integration of the consciousness of the cells (or some of them)’ that formed his body.  Recently this has been addressed by Susan Greenfield, who has proposed that short-lived coalitions of millions of neurons are responsible for consciousness.  These ‘neuronal assemblies’, which last for fractions of a second, link local events in individual cells with large scale events across the brain and many of ‘these assemblies flickering on and off somehow come together to provide a collective continuous experience of consciousness’.  In other words, our consciousness arises as an emergent behaviour of the myriad of interacting networks in our brain.  It seems no less fanciful that our individual minds networked together to generate a further level of emergent behaviour equivalent to the unified mind that Schrodinger conceived though, like Schrodinger, I find it utterly impossible to form an idea about how this might happen.

Perhaps, at some level we are creating a unified mind via the digital hive mind being formed by the digital devices to which we delegate some of the more mundane aspects of modern life [see my post entitled ‘Thinking out of the skull‘ on 18th March, 2015].  However, Greenfield worries about a very sinister potential impact of our digital devices, which is associated with the stimulation they provide to millions of the younger generation.  She thinks it could lead to small-scale neuronal assemblies becoming ‘the default setting in the consciousness of the digital native, to an extent it has never been in previous generations’.  In other words we might be losing the ability to create the emergent behaviour required for consciousness and shifting it to our digital devices.

Perhaps we are closer than we think to the vision in Maria Lassnig’s painting of the lady with her half of her brain outside her skull? [see my post entitled ‘Science fiction becomes virtual reality‘ on October 6th, 2016.

Sources:

Erwin Schrodinger, ‘Mind and Matter – the Tarner Lectures’ in What is Life?, Cambridge: Cambridge University Press, 1967.

Susan Greenfield, A day in the life of the brain: the neuroscience of consciousness from dawn to dusk, Allen Lane, 2016.

Clive Cookson, Know your own mind, FT Weekend, 15/16 October 2016, reviewing Greenfield’s book.

Nilanjana Roy ‘What it means to be human’ FT Weekend, 17/18 September 2016.

Emergent inequality

115-1547_IMGI wrote a few weeks ago about my visit to a conference on high-performance computing and big data [see ‘Mining Data‘ on February 12th, 2014].  We are able to use high performance computers to create simulations of complex engineering systems before we embark on the usual costly, and sometimes catastrophic, construction of the real system.  Some complex systems exhibit emergent behaviour, meaning that although we understand and can model the individual components when we connect them together the system behaves a new and unexpected manner, which is why it is good practice to simulate a system before building it.  Manuel Delanda has written eloquently on the topic of emergence in simulations in The Emergence of Synthetic Reason.  I encourage my first year thermodynamics students to read at least the first chapter which an amazing tour-de-force that ranges effortless from spontaneous flows of energy at the molecular level to the formation of thunderstorm systems.

Nature has many systems that could be described as emergent at some level or other.  For instance, the ants in an anthill go about their simple interactions but have no idea about how the anthill works or, perhaps more amazingly, the rafts that an ant colony can form using their bodies during a flood, as shown in recent research by Jessica Purcell and her co-workers at the University of Lausanne. With the exception of the queen, there is no leader in an anthill and all of the ants appear to be equal.  The same is not true in human society where currently 1% of the population own nearly half of the world’s wealth.

Seven out of ten people live in a country where inequality has increased in the last 30 years according to a recent Oxfam report.  This is bad news for everyone, including the wealthy because Richard Wilson and Kate Pickett have shown that in developed countries, there is a correlation between the incidences of mental illnesses and the level of income difference between the rich and poor.  A more recent study of the US found that depression was more common in states with greater income inequality, after taking account of age, income and educational differences.   Wilson and Pickett conclude that we become less nice and less happy people in more unequal societies regardless of our position in the social spectrum.

Sources:

http://in.reuters.com/article/2013/10/09/creditsuisse-wealth-idINL6N0HZ0MD20131009

http://opinionator.blogs.nytimes.com/2014/02/02/how-inequality-hollows-out-the-soul/

http://www.huffingtonpost.com/2013/11/01/income-inequality-depression_n_4190926.htm

http://www.huffingtonpost.com/winnie-byanyima/a-plan-for-tackling-inequ_b_4768096.html