September 01, 2009

Notes from interview with Donna Chung, School of Health and Social Studies

She thinks she is a foundationalist thinker, in the sense that she mainly thinks our knowledge emerges from building blocks.  Hence, she thinks the emergence of new knowledge often involves looking at data differently, or asking different questions. 

For instance, regarding domestic violence - a main area of her research – she thinks that researchers including herself tended to totalise the violent man’s identity:  they didn’t see him as anything other than a violent man.  So interviews were rather pre-empted:  women were asked about the effects of violence upon them, their children, and so forth.  Women would mainly talk about this construct of the violent man; other aspects would or might come out in small ‘asides’ that she (Donna) didn’t pick up on or pay attention to.   Whereas if you did an interview not just focused on domestic violence, and asked the woman to talk about her life and her relationship with the man, it became clear that his identity appeared as much more split to the woman than it ever was to the researcher.  The woman fundamentally saw the man differently to the researcher. 

This is an example of how emergent material comes from what she never thought to ask before:  it comes from ‘stepping outside’, looking more reflexively, rather than the data not being there.   But she thinks too that this ‘stepping outside’ can be blocked from happening, since on some level, you’re always looking to build up your theoretical understanding or perspective.  We develop this and hang on to it, and it can be threatened by looking outside the usual.  

She also thinks about emergence happening in terms of the intersection of issues.  She’s currently developing a project with a colleague that looks at women, work and chronic illness.  Chronic illness has always been understood in a quite medicalised way, in terms of a sociology of health paradigm; the ‘women in work’ issue has set quite separately from that.  So they’re interested in putting chronic illness in third place, rather than up front, and looking at how the illness gets managed.  So this is about identity again.  This isn’t dismissing the seriousness of chronic illness, but rather not identifying a person as a person with a chronic illness – it’s looking at illness as something a woman manages, and that is only one aspect of their life rather than the defining aspect. 

So new understanding – and new theoretical ways of understanding an issue – she thinks can emerge from hints in the data that aren’t picked up for a while, from looking at data differently, and also through intersection:  people from different fields will ask new questions. 

Frances:  is there a sense that things emerge ‘out there in the world’, in our social life, that don’t have to do with our perspective? 

Donna:  Perhaps.  One possible example of this that she’s currently grappling with is ICT and its impact upon sexual identity.  She did some research earlier, asking young women (adolescents) about sexual coercion by young men, and got unexpected answers – pressures for instance to engage in far more ‘explicit’ kinds of sexual acts.  It seems that relatively easy access to explicit material – far more is more easily accessible now than say twenty years ago - via ICT had really changed the things young men were asking young women to do.  So this flags up two issues relating to ICT and its impact upon sexuality and exploitation of women.  First, how ICT ‘contributes’ to sex education, in terms of how much information young people get, when they get it, and what they think is acceptable.  And second, the implications of this for young women in intimate relationships with these young men.  Young men watching a pornographic video may understand that the story in it is inaccurate, but may see the sex acts in it as realistic, or possible – in a way that they would not have seen before – and that the way women act in those films is realistic. 

So – in terms of ICT and its impact upon sexual identity:  is this a new means to the same end, or something new?  Does ICT just expand the range of ways in which women can be sexually exploited, or violence against women can be perpetrated (i.e. more of the same pressure upon women to engage in sexual activity, though involving younger women and more women because certain pressures are now more frequent)?  Or is this something beyond pressure - are certain expectations being normalised, i.e. women can be expected to engage in more ‘explicit’ types of sexual practices in a way they would not have been 20 or 30 years ago?  And if we look at adult men in existing relationships, ICT can be used to cyber-stalk, and it greatly enhances the capacity to put people under surveillance.  For instance, men can use mobile phones to check who women have been contacting and who has contacted them. 

Frances:  How would you distinguish between changes that are more of the same vs. something new? 
Donna:  That’s where we are now – wanting to undertake research in this area to address this issue. 
Frances:  And in social science terms, the development of the theory is the identification of something new?  
Donna:  Yes.

One thing Donna is also interested in re. the use of ICT is the use of a virtual self and non-virtual self.  People can create a virtual identity online [in chat rooms, etc.], which can be completely different to who they ‘truly’ are.   In domestic violence, one of the major indicators of violence is sexual jealousy – and this can extend to (in one case) a husband’s jealousy of what he thought was his wife’s virtual (i.e. online) affairs/relationships with other men, which to the husband was as bad as actual affairs.  Virtual identities are an emergent phenomenon, she thinks, in that you couldn’t do this without the technology – it’s a truly new phenomenon in social life, not an adaptation or alteration of something that already existed.

The question of whether health is in any way something that emerges, reminds her of older sociological work (Goffman) about disease as a ‘career’/identity.  She thinks sometimes that in a current wish to be inclusive, there’s a potential for researchers to define people as a service user or a carer – to totalise their identity through efforts to engage them in research.  She doesn’t want to totalise them.  But she’s also aware of how disease can be prioritised, in terms of daily routines, things people do as part of dealing with their illness – so that disease is part of a person’s identity but in ways that are normalised, so much a routine part of life that ideas aren’t even articulated.  Though this also changes over time, and with changes of circumstance, such as with move from school to employment, for instance, or with advancing years.  And she wants to explore this too in her project on women, work and chronic illness – how people accommodate their health, or not. 

There’s the example of a friend’s daughter who has a serious auto-immune condition.  She’ll train as a teacher; her health wouldn’t allow her to do a full-time job, but she’s planning to be a supply/relief teacher so that on the day, she could decide whether she’s fit to work or not.  So this is an example of plans driven by her health.  This decision about work, Donna thinks, probably has improved the daughter’s health by reducing her anxiety.  She doesn’t have to worry about her ability to manage a full-time job; she’s not fully healthy, but is active to the extent she can. 

These kinds of issues/accommodations, and the impact of illness (including changes in or onset of disease, unpredictable disease – and is there a point at which it might be economically more viable not to work, and to go on benefits instead?), at various stages of life and disease and in various life circumstances, are what Donna wants to explore in the project. 

June 15, 2009

Notes from interview with Margaret Thorogood, Warwick Medical School

When asked about how emergence applies to health, Margaret thinks of this in terms of emerging technologies, emerging knowledge, and emerging trends – for instance, recent news of the emerging trend of heart disease among young women (it’s going down in every group except women under 50).  This is one type of emergence - where trends emerge from the data.  Emerging technologies – we get a way of doing things we didn’t understand before, such as a drug or a procedure.  There’s emerging knowledge all the time.  All three of these things interact in ways that aren’t necessarily logical to affect what is policy and what is practice.  

Emerging trends – the term ‘emergence’ doesn’t get used in epidemiology, ‘emerging’ might get used perhaps, but more likely is a statement that trends have changed, for instance.  Though re. South Africa, it’s common to talk about the emerging epidemic of cardiovascular disease.  Re. this use of the concept/phrase ‘emerging’ re. South Africa, she uses it to try to get across that the trend is new, it’s a change and that it’s growing, it’s going to get bigger. 

Especially when writing for funders and policymakers, for instance, it’s important to get across that you’re ahead of the game, you’ve spotted what’s happening, and that they need to spot this too.  She uses ‘emerging’ quite frequently actually in funding applications. (Frances commented that other academics had said this too:  the term ‘emerging’ used in funding applications but not in their publications). 

Margaret doesn’t think ‘emerging’/’emergent’ is a scientific term and hence she wouldn’t use it for (other) scientists.  She may describe something as ‘emerging’ in a funding application, but if challenged, she would have to say that she thinks this is a new trend but they don’t have the data to prove it absolutely, since for instance they do not have 200 years of data about cardiovascular death rates in Africa.  So the ‘emerging’ here refers to something new – there is no record of cardiovascular disease in Africa 200 or even 50 years ago.  In the UK, the 19th and 20th centuries saw a huge rise in cardiovascular disease, but it isn’t known how much this relates to doctors’ ability to diagnose it or changes in the recording of it (it is known that death certification changed enormously in that time).  You wouldn’t talk now about an emerging Aids epidemic because this is present and known – and hopefully rates are falling rather than rising.  So for her, ‘emerging’ is something that’s coming out:  it’s new or it’s growing.

Why is cardiovascular disease rising in South Africa?  We don’t really know.  Hypotheses include the salt hypothesis (i.e. salt used more because it’s more available) although Margaret doesn’t believe this.  Barker hypothesis:  people now in middle age were born in an age of extreme poverty to underweight mothers.  So in utero these people’s cardiovascular systems were ‘programmed’ for famine, and are now being put under enormous pressure by growing body size reflecting more plentiful food.  And there’s the hypothesis of Westernisation, with more smoking and drinking and a higher fat diet.  It’s very difficult to sort these out when there’s a very sparse historical record (such as the odd missionary record). 

Back to the issue of women under 50 dying of heart disease (BMC paper) – hypothesis is that this is largely due to obesity, less exercise, more smoking.  Margaret wonders if country of birth has been taken into account in the paper:  this trend (i.e. death rates not going down) might be due to changes in the British population, namely immigration from South Asia.  If country of birth has been taken into account, then we need to go back and look at factors at play – the known ones (such as body mass), also some that might not have been taken into account, such as contraception (i.e. birth control pills), which are said to increase the risk of heart disease very slightly, but what if the risk isn’t so slight?  So this is a case of emergence where much might be explained by going through what we already know. 

Frances:   is there a case of emergence where we can’t explain it, even though we have a lot of information about it? 

Margaret:  Yes, two trends.  1)  The rapid increase in childhood asthma – not yet explained.  Nor is this an instance of emergence (potentially) due to change in (the boundaries of) diagnosis, as she thinks is the case with the increase in childhood autism.  2)  Increase in peanut allergy similar – not explained.  Neither increase can be genetic, because it’s happened too fast.  (As opposed to insulin-dependent diabetes, which is increasing in children – this might be selective survival, since Type 1 diabetics started surviving and hence reproducing, whereas earlier they would have died without having had children.)  Peanut allergy must be an environmental change, but not explained.  First you ask, is the increase real?  The answer seems to be – yes, at least some of it is; and yet peanut allergy is virtually unknown in other countries, such as Israel and even France.  Nor is it a matter of definition – observed asthma and allergic reactions to peanuts are pretty straightforward. 

Frances:  are there situations where we can’t explain the change/trend? 

Margaret: Yes, loads of situations.  Most of the time, we just don’t have the data to examine.  It’s a philosophical point of view as to whether you think ultimately there is an explanation for everything (which is her viewpoint), if you understood enough; or whether some things just happen and are simply inexplicable.

Frances:  There might be an in between, where some things are explicable by 2-3, perhaps 4 major determinants; but if there are many different things interacting to produce the change/trend, at the moment anyway we would find this very difficult to understand, even if we had the data. 

Margaret:  Though it’s very difficult to unpick what we mean there by having the data. 

Agreed by both – we have to know what to look for in order to have the data. 

Margaret:  And people are too complicated to ever have perfect data.  And plenty of other things have been studied but still aren’t understood, for instance re. multiple sclerosis:  huge amounts of work have gone into finding the cause for this, and we know some things about it (for instance, more common in colder northern and southern hemispheres; it involves an autoimmune response), but we still don’t really understand what causes it. 

There had been thinking, at least among some epidemiologists, that we’d conquered infectious disease – but then along came the AIDS epidemic, and SARS, and drug-resistant TB...  All of these made epidemiologists, and others, realise that we are never going to conquer infectious disease, and there will always be a new one or the disease changes.  The situation is not static

Margaret’s brother suggested as children born just after the war, adults around them ‘conspired’ to say the things were permanent – because these adults had just been through the most enormous, horrendous upheaval.  So now, as an adult, her brother and she and others like her have had to adapt to the idea that things aren’t permanent, and in fact are changing all the time.  Whereas their children just take for granted that things constantly change.  Same thing for infectious disease – after the war, ‘the idea was that with peace, we were going to conquer everything.’  And they kind of forgot that things change and the unexpected can happen.  Margaret has not seen this kind of issue written up anywhere – though Martin McKee (School of Hygiene) has talked about this somewhat, about how governments for instance assume that diseases can’t cross the animal-human barrier.  Also Tony McMichael (formerly School of Hygiene, now back in Australia), a global change person:  also saying ‘look behind you’, for instance warning that malaria will be coming back to England. 

Another driver of emergence:  medical technology, and technology in general which is a huge driver in things like food change.  When they worked out how to skim milk efficiently, skim milk stopped being more expensive than full fat milk and sales took off (skimmed milk lasts longer than full fat). 

January 16, 2009

Emergence and health

Frances Griffiths

Health can be considered an emergent phenomenon for a society, community, family, individual, cell, and arguably lower levels of analysis such as enzyme cycles or genes. By considering it an emergent phenomenon I mean that it is not possible to fully understand how it comes about. For example, a biochemist is able to specify the structure and mechanism of two, three or perhaps four molecules interacting within a cell. One account of the whole cell is that, if we knew enough about all the molecules and their interaction within a cell we would be able to explain the state of the cell as a whole. The complexity account of a whole cell is that there are attributes of the cell as a whole that will not be fully explained by understanding all the molecular interactions within the cell, just as temperature cannot be fully explained through knowing the details of molecules.

Health at one level of analysis does not seem to determine health at higher levels of analysis. In gene therapy, where the health intervention is targeted at a specific gene of the individual, but the same result is not produced every time as the effect is modulated by the complex systems of cell, body and so on. In epidemiology, the determinants of health on the individual level are not the same as those on the population level.

Health can remain in some sense stable – an individual remains healthy or not, a community may have a consistent expectation of life span and wellbeing for its members. There may be variation within this relatively stable state but it remains recognisably the same. At times health may change dramatically – cells become cancerous, an individual becomes disabled or terminally ill, a community may be decimated by epidemic.

The history of the cell, individual or community impacts on its emergent health and the effect of interventions aimed at improving health. Two individuals who have apparently similar health issues may respond very differently to an intervention, be it gene therapy, a drug, surgery, or behavioural intervention because they have different history and so different ‘learning’ or ‘adaptation’. This is so for all plants and animals.

Health care interventions, which are often based on our understanding of two or three processes that lead to ill health, may work as intended but can also seem not to work perhaps because we are missing something in our evaluation or there are unintended consequences that our understanding of the health problem could not predict, or there may be compensatory mechanisms working against the intervention that we were not aware of. These mechanisms may be quite different from the focus of our intervention, for example social interaction or social policy may modulate the impact of infection control at the biological level.

September 02, 2008

Sarah Hodges, History

Trained as a Modern South Asian historian, studied in the U.S. in ‘postwar area studies’ paradigm.  The U.S. had to create expert knowledge about ‘strategic’ other parts of the world, including South Asia. She had very interdisciplinary training, especially in history, anthropology and area studies [languages and cultures of a particular area], with language training in Tamil.  She’s most comfortable academically in the field of history, but has tended to pursue topics that aren’t traditional for the field of history because they are contemporary, and/or ‘useful’ topics.  So her PhD was on the history of birth control in India – she addressed health and governance under colonial rule, and the ‘space’ of the colonial subject. 

Her current project is about clinical/biomedical waste [surgical gloves, linen, IV tubing, etc.], globally and especially in India.  This also addresses the relationship between health and governance, because health is both an object of governance [example: immunisations] and a mode of governance [example:  epidemiology].  The regulatory apparatus to address biomedical waste has focused on infective waste.  But now, in the era of globalisation, the policy framework needs to address commerce:  there’s a growing industry to recover supposedly disposable biomedical waste that is being repackaged and resold.  So there’s a tension between current regulation and the on-the-ground reality.  Is biomedical waste regulated as an object or problem of risk, and/or as an object of commerce/trade?  She’s looking at value chains:  where does the waste go, who buys it, who sells it, when does it acquire value [in both emotive and economic ways]. 

Medical tourism is huge issue within India.  Hospitals for tourists [5 star] are ones who have been most compliant with legislation to dispose of their biomedical waste separately from their general waste – they’re the only ones who can afford to do this.  Because there’s still so much value still to be extracted from this biomedical waste, especially plastics [blood bags, tubing, IV catheters, etc.], it regularly ‘disappears’ from the disposal process.  There’s always been a ‘second sale’ market - you get hospital orderlies [low paid, low status] who pocket say syringes and then sell them on – but that’s ‘small potatoes’.  It’s the big, tourist hospitals that that offer maximum opportunities for securing supposedly disposable material that is of maximum value.  The conjunction of biomedical waste and medical tourism is an unanticipated, unintended outcome of this global India. 

It’s urgent to inject greater analytical framework into how the field of history of medicine has been carried out to date, including for addressing issue of bioethics.  Bioethics (in many forms, from medical school curricula to media) has become incredibly important – one problem is radical ahistoricity, and privileging of the ‘lay’ attitude as if it were some kind of free-floating entity.  It’s important to ask bioethics to reclaim some politically/ popularly relevant questions and make us rehearse how we got there.  Much bioethical discussion now, she thinks, serves to legitimate contemporary genomic research and practice, rather than asking critical questions about ‘truth claims’.  When you talk about history, you can’t ignore political questions.

The field of the history of medicine has largely organised itself largely the way that medicine has:  in terms of diseases, techniques, technologies, specialties.  The questions that need to be asked now are more reflexive, examining implications of work being done.  For example, a history of diabetes could be placed within a number of broader conceptual frameworks, such as the history of increasing body mass, colonialism, changes in lifestyle, globalisation, etc.  There are different ‘wagons’ to hitch the ‘what happened’ narratives to.  Historians are very good at putting together ‘what happened’ narratives – they know where to look things up, etc. and they know how to recognise similarities and differences over time.  But in the field of the history of medicine, the history has not always been considered within a context of broader social and political processes, which would provide some explanatory weight.   

Idea of emergence:  poststructuralism has encouraged historians to ask questions about how objects become objects – including how they become fit objects for investigation.  So this is an instance of emergence.  One main way of discussing emergence in contemporary writing especially about the developing world has been through a language of ‘legibility’ – the term comes from a book by James Scott.  As a marker of being hip and up to date with new work in the social sciences, especially if working on Asia or Africa, one would speak of making something ‘legible’.  Legibility here means emerging as an object of knowledge.  Involves a spatial argument:  that colonial regimes in places like India and Africa could only operate given their incommensurability – their cultural, linguistic, etc. cognitive dissonance.

Re. patterns and emergence [for example, patterns of contraception use in colonial India could be viewed as an emergent phenomenon within the population]:  there’s a straightforward temporal explanation.  What she thinks is more interesting is how things become noticed:  this is what she thinks of as emergence.  Legibility:  you have to notice things, and then put them in some kind of framework.  And things that aren’t on the radar, so to speak, are illegible and therefore can never emerge.  Illegible things are the ‘unknown unknowns’ - you can’t even imagine what they might be. 

[[ Mathew Thomson (also in History) - does work on mental illness.  He thinks we’re now past the time when historians were confident about looking to particular places for explanations [with political historians looking to politics, etc].  There’s been synergy among different levels; it’s become hard not to be aware that you’re missing part of the answer unless you look elsewhere.  Historians’ answers have become very complex, and trying to explain why certain things emerged at a certain moment usually involves lots of things happening at the same time.  Tricky to give clear answers.

Example - mental deficiency:  He’s interested in how a particular category [mental deficiency] emerged at a particular moment, and was seen as a problem.  His work draws upon medical literature and psychological literature, and includes involves thinking about political ideology, institutional systems including the Poor Law and welfare provision and the education system, structure/forms of the family and how the family was changing.  This involves multiple series of explanations that come together and are connected in a complicated way. Emergence in this sense is a fundamental historical question:  why something is the way it is, at that time.  ]]

Legibility addresses how things come to be noticed and defined.  Frances asked:  is there a sense in which things may be there as an emergent phenomenon, but not noticed?  Sarah thinks yes.  But that’s not the historian’s job – history looks at the various forces that have an influence.  Dominant explanations/explanatory models – what’s assumed to be the explanation for something – need to be interrogated.  For example, when doing her work on the history of birth control in India, people assumed ahistorically that there was no history of birth control in India – that India has always been overpopulated.  The dominant explanatory model of population control is that people are ‘poor, lazy and stupid’ – but this is, of course, laden with assumptions.  Re. India, this explanation is the product of a particular historical moment and set of circumstances, and an explanation that she rejects.  Nonetheless, she does think that even if something hasn’t been made legible, there is a realist sense that the something is there – that the act of representing the past, in and of itself, isn’t what determines whether something existed. 

But one is always writing about the past through the filter of the present (which is a basic poststructuralist idea).  Emergence involves how one represents the past, and the only tools to understand, or represent / make visible, the past are the tools of the present.  Emergence also involves conjunctions:  something can only emerge in conjunction with other things, and historians are necessarily predisposed to multicausal explanations of why things emerge. 

[Frances noted that most interviewees have discussed 2 main ideas or uses or ‘emergence’:  

- 1 to do with lots of things happening, and then something happens [emerges]
- 1 about something becoming visible, sometimes described in terms of surprise – the person hadn’t seen it before, or didn’t expect to see it.]

August 06, 2008

Notes from interview with Simon Bright, HRI

[Frances asked about issues of food supply/price.  This goes back to the 1960s, and the push for higher food production and food security.  The ‘green revolution’ was high-tech; it had social consequences as well – some people were winners, some losers.  Industrialisation of agriculture led to huge price drops in food staples.  In last 3 years, prices in real dollar terms have increased approximately 2.5 fold.  Drivers:  world reserves of food have dwindled; supply lowered through some bad harvests; emergence of China and India as big industrialised nations, with high demand for meat and therefore grain; emergence of biofuels driven by subsidies.  Knock-on effects include in the UK:  no payments for set-aside land here, hence more land being brought back into cultivation.  Food production brings in all kinds of other issues:  issues of environmental footprint [especially comparing nationally produced with imported food]; organic food, tied up with attacks upon cheap food production; FairTrade  which touches upon international development issues.] 

Challenge of delivering sustainable development, just like sustainable healthcare:  how do you achieve this in a consumer society where there is lots of free choice?  Government doesn’t like having to tell people what to do or how to live, and it’s very difficult to achieve ends in this way in any case. 

What emergence means to him in his disciplinary framework:  in systems biology, there’s a reductionist perspective - one thing you can do to understand a system is to look at all the parts.  Emergence in this sense:  you start with all the genes, and you end up with a living organism.  There’s a belief that one can explain the properties of an organism by the properties of what’s in it and how it’s organised.  This compared to the ‘holistic perspective’:  I need to understand the whole thing and then I can understand how it works. 

He thinks the reductionists are mostly making more progress than the holists, because the reductionists can work away at collecting all the parts [like sequencing genomes], even though they haven’t really explained how you go from genes to the organism itself.  He can model what’s going on in a system, and one can build up to the ‘next level’ system from smaller elements [such as how parts of an organism work, for instance]. 

He thinks of emergent properties as how the responses / properties of the whole organism emerge from the underlying smaller elements or ‘modules’.  There’s a big research agenda now in the international community:  we think we have enough tools now [with modelling, gene sequencing, etc.], and a big enough parts list, that we should be able to say how a plant works – enough to construct an ‘e-plant’. 

Another approach:  someone is studying how you get from the behaviour of individual insects (here, ladybirds) to how communities of them work in a certain environment (leaf canopy), including in their interaction with other communities (ladybirds and aphids).  This work might then be able to predict interaction in another environment.

Also – the issue of scale.  Example:  diffuse pollution.  A team has measured how much nitrate is in a river.  This one team started measuring every 10 seconds; they could separate long-term (hours/days) and short-term (minutes) activities.  They discovered that it looks as if different things control the short-range and long-range – it’s not a summation, the short-term elements don’t just add up to the long-term.  Some things only happen on a certain scale.  The smaller elements must interact in some way to give you the end product - but how is not clear.  This is where complexity comes in:  you try to find out what you need to focus on in order to know certain things.  New measurement technology and mathematical tools allow you to approach problems in more detail and in new ways, because you can you can really tell what’s going on, rather than having to make assumptions.  [Simon can send Frances the slides]

Challenge:  they [at HRI] tend to look at individual plants or communities in a field.  But sustainability he thinks is likely to be determined on the scale of a landscape – so again, issue of scale. 

Re. modelling:  he thinks there are 2 sorts of models, depending on what happens at the ‘boundary’.  One sort models and describes what’s in an area, but doesn’t address drivers/causes and doesn’t make any predictions about what’s happening outside the boundary of exactly what’s being modelled.  But a model based on reality should be able to make predictions about how something would work ‘out there’, beyond the boundary – predictions that make some sense and that should be interesting rather than just trivial.   

Circadian clock [David Rand]:  model developed of how the clock works, based on 4-5 genes.  The model made a prediction about what another gene should do, although this was true only in a mutant.  David’s idea:  elements of a clock should be the same regardless of the period being measured [a season vs. a day, for example].  Next phase would be synthetic biology, where you use your tools to try to create an organism that behaves/responds in a certain way.  This likely to be a hot topic over next 5-10 years, especially for microbes because they have far fewer genes. 

Re. plants, you can currently add and take away functions – but not yet start from nothing, so to speak, though in principle he thinks this is possible:  to build up a plant from smaller elements (the ‘modules’), without there being the need for anything extra (some extra principle). 

What about the impact of smaller variations when you’re putting together a plant?   He says that when Dolly the sheep was cloned, it was clear that there were lots of elements that weren’t understood.  And if you’re doing the same process 10 times, you don’t necessarily get the same results 10 times.  Small changes do make a difference.  Relevant to gene therapy in humans, this suggests that some processes won’t work every time - a problem when each individual is important. 

Also, limitations in introducing change:  their aim [in a former company, Zeneca] was to make a tomato with ten times the lycopene (for health benefits).  But it didn’t work.  They could get up to 2.5 times the lycopene by changing all the genes in the pathway, whereas in rice, with the same genes, it was possible to increase the betacarotene (related to lycopene) by 1,000 times+ [for golden rice], from a very low base.  (There are probably other elements involved: in a tomato, lycopene is stored in/under the skin – too much lycopene would just clog the skin.)  With the tomato experiment, this was done using a stepped, ‘Lego approach’ (reductionist, in the terms used above) to try to achieve change, rather than a full understanding of all the changes involved in intermediary steps (which would be a ‘holistic’ approach).

Frances asked re. the ‘Lego’ model of biology/biochemistry:  is time involved?  He thinks it is, along with scale.  Re. biological processes involving time, such as plants that need a cold season before they flower:  there’s just beginning to be an understanding of the epigenetic biological processes at molecular level involved here [i.e. it isn’t really just a Lego brick, it might depend on the history of the brick ...]. 

Systems biology at the moment has constructed itself at the cellular level and below, because this is the level at which the biologists think they can get enough data.  Epidemiologists and ecologists have lots of data but far too many variables to be able to predict things / work out causal models (at the moment).  So it’s a matter of where you can get enough data.  There’s no particular reason why you should model at the lowest level – it’s just that’s where the data is (and presumably also the tools to deal with that data) at the moment. 

July 23, 2008

Follow–up to interview with Graham Medley

Follow-up to Graham Medley from Understanding emergence and evolution of health

For an interesting discussion of shifts in epidemiology, and an argument for the need for the field to move away from a focus upon risk factors (for non-communicable disease) to a broader examination of "the pathways by which biological and social experiences generate health and disease, and the impact of biological and social changes on the health of populations", Graham suggested:

March D, Susser E.The eco- in eco-epidemiology. International Journal of Epidemiology 2006; 35(6):1379-1383

[From within Warwick, if you're signed in, you should be able to get access to the article by clicking on the link.  If problems, please email Janice on]

May 05, 2008

Reflection after interview with Alison Rodger

Follow-up to Alison Rodger from Understanding emergence and evolution of health

What is the degree of variation between things that are the ‘same thing’ in different disciplines (e.g. molecules, animals, people). Is any apparent lack of variation because of our limited understanding/techniques or is it ‘real’?

Alison Rodger

Alison Rodger, MOAC

Notes on interview, April 2008

Emergence: when something happens that we don’t expect from what we know of the behaviour of the components.

A complexity term

Biochemists: if something different happens that we don’t understand, then usually there are variables we haven’t controlled or don’t know about. We usually eventually find it.

Example: peptide from the venum of the bee sting. When it inserts into the membrane of a cell, it disrupts the cell membrane. Experiments showed variable (often no) results. The important variable was time – found by chance. The ‘signal’ increased then decayed over time, so most of the experiments missed the signal. Time was important in terms of when the signal started but duration of the signal was also important (the latter was what was missed initially).

Systems such as cells are reduced to models of the system - molecular models. Limited number of components so it is usually possible to work out what the variable is that has been missed. For example, cell membrane in a buffer. Current technology/techniques require simple models (limited number of molecules in the model) otherwise the ‘signal’ is so muddled it cannot be decomposed (to understand what is happening).

We are limited to understanding 2, 3 or 4 body problems. Most experiments are looking at two molecules. Some look at three e.g. need three molecules binding for anything to happen – and they may need to bind in a particular order before the expected effect happens. Example DNA + enzyme + drug that inhibits enzyme. DNA + enzyme have to bind before the drug has an effect.

(Example of an Enzyme is topoisomerase – that changes DNA coiling).

This is more complicated than past experimental technique has been able to show (not complex). Experiments can be done on real cells but the experiments are ‘crude’ e.g. does it kill a cell or not, does it get into a cell or not. Limited by our techniques. We may be able to show a molecule gets into a cell, binds to DNA and stops replication and cell dies. However, in reality the molecule binds to all sorts of things. If we had more information, such as what the molecule binds to, we would know more and so be able to control and intervene more. However, this information may not help with the problem we were investigating.

It may be or should be possible to build up from simple biological models about what molecules do to understand a real system such as a cell, but we are a long way off that at present.

Mark Rodger: uses classical mechanics to develop simulations of how molecules are interacting with each other. Assumption of similarity of molecules.  His research group:

Biochemists think in terms of molecules.

Biologists may think in terms of function.

Function may be considered emergent, as it can be understood (to some degree) without having to add together component parts to explain it. However, if we understood enough, then we could explain function in terms of biochemistry and so it would no longer be emergent. Biologists may not need to know about molecules but they can make big errors if they don’t (e.g. some drugs, once the biochemistry is understood it can be obvious they are not safe).

The orientation of a molecule can make a difference to what it binds to and thus the effects it has. If a molecule (such as a drug) does not have the right orientation, then it may start a cascade of effects leading to a problem. Sometimes the consequences are obvious, such as if a drug binds to DNA and cleaves its backbone, this part of the DNA cannot replicate and so a protein will then not be produced properly. Currently there is a limit to how far the effect can be tracked of a molecule with the wrong orientation – may not explain symptoms or death of organism. However, it is all inter-molecule interactions so it could potentially be understood at this level.

Molecular system very complicated in a finely balanced system such as a cell. There is duplication of function and redundancy. Two molecules may do the same thing so one can take over if the other one is knocked out. Also a lot of feedback and compensating systems. These may look complex but this is only because we don’t yet understand them.

Observation (e.g. anti inflammatory effect of methotrexate) can be useful for noticing things we don’t expect – because we don’t know enough about the molecules.

Graham Medley

Graham Medley, Biological Sciences, University of Warwick

Notes from interview April 2008

Emergence is a property of complex dynamic systems. The interaction of the different components and processes give properties of the system that are not those of the components and processes.

Example of system: a population of individuals where the state of the population is determined by the state of the individuals and the state of the individuals is determined by the state of the population. Where there is this type of feedback there can be emergent properties. With this definition, an epidemic is an emergent phenomenon.

The challenge for understanding complex / complicated systems is identifying the key issues, i.e. variables and outcomes. Most natural scientists believe there are only two or three important processes that explain a(n) (emergent) phenomenon, we just have to find them. Ultimately, demonstration of understanding, and usually the process of developing the understanding, comes through experiment.

Historically, when a phenomenon “emerges”, it is possible to look back and work out what was happening. We can learn from this experience but we may still not be right next time, suggesting that our explanation was wrong, or that the same phenomenon can have multiple (>1) causes.

Health interventions frequently do not have the desired effect:

-  we are missing something (e.g. we are assessing the wrong thing)
-  interventions can have unforeseen/unanticipated consequences
-  often get “compensatory” mechanisms that work against an intervention, i.e. reducing one risk results in people changing behaviour to increase other risks.

Complicated systems – lots of variables – can have emergent behaviour – can be complex or “straightforward” (e.g. linear).

Complex systems – different levels of interaction – can be complicated or simple.

Complex systems have non-linear dynamics – may be multiple catastrophes, multiple stable states etc.

All real systems are complicated. They may not be complex (defined as above). We divide our complicated world into chunks to understand it, and may loose essential aspects of the whole system (if it is complex) by doing so. Human systems are often (always?) complicated. There is lots of compensation going on to keep them stable; they are not necessarily ‘complex’ (as defined above), and are probably less complex than they should be.  See Cohen J and Stewart I, Collapse of Chaos

Emergence has two meanings when applied to observations about system behaviour:

- complicated, lots happens, didn’t spot what it was that happened bringing about outcome. Can often only spot in retrospect. Usually means that we were missing something important. Usually only spot the emergence and then what was happening when something changes, something emerges. Difficult to spot something that is already there as we don’t see it emerge.

- emergence even when we understand what is going on: e.g. the NHS (how did it come about; could it have been predicted); would the French army have marched on Moscow even with a different leader (e.g. had Napoleon died in childhood).

Example of research undertaken and published by GM:

Hepatitis B

Key issues about hepatitis B: the younger an individual is when infected, the more likely they are to become a carrier. The more carriers there are in the population, the younger individuals are likely to be when they become infected. The result is populations with two possible stable states: either a high level of carriers in population or a low level of carriers. Only get an in-between state if a population is changing from one stable state to another. (The influence of age on the development of Hepatitis-B carrier state. Proceedings of the Royal Society of London, Series B, 253, 197-201, 1993. Medley, G.F., Lindop, N. A., Edmunds, W.J. & Nokes, D.J. (2001) Explaining HBV endemicity: heterogeneity, catastrophic dynamics and control. Nature Medicine 7, 619-624)

Bovine TB (bTB)

Epidemiology and changing behaviour of farmers help to keep some infections endemic, including bTB. Farmers have to learn to live with it, running their farms around the bTB tests. Most rural vet practices survive because of the payments they receive for conducting TB testing. Also takes vets out to see livestock. These benefits would be lost if TB was eliminated. Paying compensation to farmers for infected cattle is also somewhat contradictory, in that it essentially rewards farmers for having infection, however it is necessary to provide compensation to farmers for infected animals otherwise they may hide them.  Compensation was an essential component of BSE control.

Foot and mouth disease is treated very differently – precedents were set in the late 1800’s and reinforced in the 1960s when it was stamped out. However, given the outcry with the last episode, whether it can be stamped out with future epidemics is unclear.

GOLD ( Governance of Livestock Disease (project in the Rural Economy and Land Use programme). This in an interdisciplinary project (epidemiology, politics, economics & law), aimed at understanding how control policies are developed and implemented. We will be studying 5-6 diseases simultaneously, which is itself unusual as most policy and research focus on one disease at a time. Control of disease is a multilevel issues: governments make policies (largely as a result of international trade agreements) and farmers make decisions (some of which depend on what other farmers are doing, so that they are actually playing games).

Somatic cell count (SCC) in milk (essentially a measure of the density of white blood cells). Cows with intra-mammary infections (mastitis) have high SSC. Farmers are paid a premium for a low SCC, and are unable to sell the milk when cow is being treated with antibiotics. Consequently, farmers try to reduce SCC and incidence of mastitis. But a very low SCC appears to be related to an increased risk of acute (and possibly fatal) mastitis. Although this is still contested in the literature it suggests that there is an intermediate optimum.  (See in particular: Preventive Veterinary Medicine 64, 157-174, 2004.  May also be of interest:  Journal of Dairy Science 91, 1403-1415, 2008; Journal of Dairy Science 87, 1256-1264, 2004.)

notes after interview

Follow-up to Michael Parchman from Understanding emergence and evolution of health

Frances Griffiths notes after interview:

MP has instigated and evaluated successful practice based changes in relation to diabetes care: published.....

Veterans Health Administration (a major health care provider in the US providing health care to all Veterans) describe a muli-level approach to smoking cessation (effect not yet fully evident) in BMJ 2008; 336:1016-9

Sign In

To see blog entries, click the ‘Sign in’ link at the top right of the screen

Search this blog

Blog archive

Not signed in
Sign in

Powered by BlogBuilder