I’ve written a couple of posts recently (here and here) about how the AGW scare has been picked up as a variant of religion, and defended by scientists who refuse to make their data and code available to others. In this post I want to look at a common intellectual game in which someone who has a hypothesis protects it against attack by developing another hypothesis that purports to dispose of the first attack, so that no concession has to be made about the original hypothesis.

You can see a lot of this in the AGW issue, such as the hypotheses that have been brought in to explain why there is an apparent cessation in warming, despite the steady increase in carbon dioxide accumulations in the atmosphere. But I thought we might give AGW a bit of a rest, and look at something else that was very much part of my life thirty and more years ago: the relationship between class and party. You will see that there is a strong resemblance in my story to AGW.

In the 1960s and 1970s the consensus was that Australian politics was really the politics of some kind of class struggle, in which the workers fought the bosses through the Labor-versus-Liberal electoral battle. Indeed, some writers saw no difference at all, and cited opinion poll data in support. I had puzzled over this equation as a young man, because it didn’t fit with my own experience, or that of my family back into the 19th century. And in fact, if ‘working class’ and ‘Labor’ were truly equivalent, then all working class people would have voted Labor.

But they didn’t. In the rigorous survey work I did from the late 1960s to the late 1970s the ratio was about 1:2 — that is, about a third of the workers voted for the Liberal-Country Party coalition of that time, and two-thirds for Labor. Much the same proportion was true on the other side — about a third of middle-class people voted Labor.* The more I investigated, the less powerful the thing called ‘class’ seemed to be as an explanatory variable in Australian politics. People knew what ‘class’ meant, and could locate themselves in class terms, but the term had no real emotional force .

At the time I was writing, another Australian political scientist, David Kemp, was coming to the same conclusion from his own detailed study of Australian public opinion data. While I was a country chap with no connections in politics, David Kemp was the son of  Charles Kemp, who was the first Executive Director of the Institute of Public Affairs. The two of us had at that time had little connection (he was in Melbourne and I in Canberra and then Sydney), but before long we seemed to be cast as the image-breakers of what had been settled science in Australian politics and history. David Kemp went on to be Malcolm Fraser’s chief of staff, then a Liberal MP and Minister in the Howard Government.

Two books re-asserting the importance of class came out in the late 1970s, not exactly in response, since they had been in preparation for some time. They were the work of two most able political scientists, R. W. Connell and T. H. Irving. Neither engaged with survey data at all, since ‘class analysis’ starts with theory and ends with theory (or for some, with ‘praxis’). But at conferences and seminars in the mid to late 1970s the class/party issue was discussed again and again, theorists against the empiricists.

For Connell and Irving, and those who agreed with them, the survey data were irrelevant, but since the survey data were, at least on face value, discordant with the notion that there was a ‘ruling class’ with a ‘ruling culture’, the theorists agreed that the people we interviewed must have had ‘false consciousness’, a term coined by Engels and revived after the Second World War by Herbert Marcuse. People who had this particular disability did not recognise their real situation in the class war, and that seemed to mean that most of the random sample of 2054 Australians we had interviewed in 1967 were disabled in this way. Maybe all Australians, all people everywhere, had a false consciousness about who they were and what they were doing, and not just with respect to’ class’.

A proposition of this kind is a debate-stopper. If people could display false consciousness, then empirical data about what they themselves thought and felt about things were beside the point. The Theorist alone could see what was happening, because he or she understood the Reality. Understandably, I felt rather peeved about being told that my years of work and study were pointless but, as it happened, ‘class analysis’ had a brief moment of glory in Australia and then faded, while survey research keeps on going, and for my part I was soon to leave that field of work for policy and administration.

Since then I have seen the same kind of  intellectual coup d’état used in other debates from time to time, and find myself on the side, always, of those who want to inspect the data, rather than just listen to the Theorist. It’s not that I am opposed to theory, not at all — we advance when new theories are put forward. But we advance fastest when the new theory is put to the test.

[None of my work of that time is online, but some of the story and data can be found in my Stability and Change in Australian Politics (second edition 1982)]

Join the discussion 20 Comments

  • Peter Kemmis says:

    Were Connell and Irving strong supporters of the “working class” (those who worked with their hands and bodies as well as minds), and/or of the left in politics?

    • Don Aitkin says:

      They were certainly of the left in politics (you would not write books like these and be a ‘conservative’), but as to their participation in ‘working class’ activities, I can’t say from my own knowledge.

      • Peter Kemmis says:

        I find it quite understandable that with all of its idealism, the left works from theory. It tends to develop ideas of how the world should be operating and overcoming its current weaknesses, of which without doubt there are many. So it adopts a top-down approach, and that’s why the left seeks to introduce considerable change through persuasion, legislation, and regulation. Conservatives tend not to seek as much top-down change, but let it happen more gradually.

        Often the left’s approach can be helpful, especially where the conservative side has become moribund or reactionary. Regardless, theory that remains unsupported by evidence, remains untested theory. At present in the West, I think we are seeing a very large and widening gap between both sides, in part because there are now large numbers of articulate and educated people in positions of influence, so the voice of the left has become very strong. Also, the greater the divergence between evidence and theory, the louder the defence of he theory.

    • Walter Starck says:

      The great virtue of theory is that it always tidy and can be depended upon to confirm what we wish to believe. Data is too often messy and has no respect for even the most fervent belief.

      Theories rendered in computer code become “models”. They embellish theory with the fail safe certainty of high-tech electronic computation. As “intellectual property” they confer the added merit of exemption from independent examination for reasons of “confidentiality”.

      The preference for theory and models and the unpopularity of data in the social “sciences” is because adherence to the latter would require discarding too much of the former. Too much of institutions, egos, reputations and beliefs is invested in theory to let mere data have such influence.

  • Alister McFarquhar says:

    I grew up preoccupied with truth as a religious concept

    And Scottish education whether driven by the Enlightenment or in spite of it was largely a list of facts

    It never struck me economic theory was anything other than a model to facilitate thought on issues

    until I was confronted with theory in the Economics faculty In Cambridge

    In the 60s I was told noone got a Ph D except on theory-and few on theory

    Very little data existed except for local case studies

    until macro data grew with the development of National statistics

    but economists have never resolved the aggregation problem

    Recently I find myself closer to the notion that religion dominates evidence

    So just put the theory to the electorate decide

    More likely to be better than deciding on basis of data that are incomplete and fiddled to suit the politics of the issue

    I make exception for the controlled experiment

    but how is this to be applied in an uncontrolled reality

    Too brief I know but enough to solicit disagreement


    • Don Aitkin says:

      It’s true that good data on a national scale are relatively recent, though Australia started well with Timothy Coghlan at the beginning of the 20th century. Good survey data have existed for seventy years now, and satellite data covering the globe began in 1979. So we are moving (day by day!) into a world where there are better and more comprehensive data.

      One problem is that we don’t live in the globe or the nation but in a local settlement of some kind, and it is the reality there that we confront on a daily basis. And while I think the ‘nation state’ is the best mechanism yet devised to bring a decent standard of living to most people, it comes with its own tensions and inefficiencies.

  • Gus says:

    Eventually, Labor lost and the Coalition won, in states and federally, after years and years of Hawke-Keating government, undoubtedly because “false consciousness” spread like flu amongst working classes. Or perhaps people no longer wanted to identify themselves as “working class” any more–is this the same as “false consciousness?” And even assuming I am a working class person and see myself as such, seeing how my own side of politics corrupts everything and is clearly no longer capable of delivering good governance, why shouldn’t I cast my vote for the opposition?
    Yes, some academics just love to segregate people and slot them into separate drawers: working classes to always vote Left, factory owners to always vote Right. But what about small business owners and self-employed?
    I also saw the Australian society divided in different ways. At the time about 40% of Australians were Irish, 40% English/Scottish and 20% others, including about 10% of mixed descent with one parent either Irish or English/Scottish. This was not class division. It was national and religious. It overlapped with class division but not precisely. As there were rich Irish Catholic business owners, there were also poor English Protestant laborers. Did they vote along national/religious or class lines?

  • PeterE says:

    Yes, I think you have found a useful parallel between the purveyors of class consciousness and false consciousness and similar concepts with the CAGW theorists. Neither seem willing to debate the issue or accept a workable compromise. This is not altogether surprising because both seem to me to belong to the same politico-religious grouping.
    Thanks also to Gus, although a report by the demographer Charles (Grenfell-) Price around 1999 suggested (based on his analysis of, I think, grandparents’ origins) an ‘ethnic weight’ in the Australian population of (something like) 49% English; 12% Irish, 11% Scottish, 3% other ‘British’; 6% German; 4% Chinese; 3% Italian and Greek and then smaller percentages for just about anyone else you can think of. This is just from memory but would be a useful starting point for determining the actual ethnic origins of Australians.

    • Gus says:

      These numbers will vary depending on whether you want both parents of English descent to call yourself English or if one will suffice. From what I remember, the number of the Irish (or people calling themselves such) was higher. But people, indeed, do mix, and this includes English-Irish couples; what do you call their children then? Or, more appropriately, what do they call themselves?
      The current fountain of all knowledge, the Wikipedia, states that
      36.1% are English
      35.4% are Australians (what does this mean?)
      10.4% are Irish
      8.9% are Scottish
      4.6% are Italian
      4.5% are German
      4.3% are Chinese
      Then you get Indian, Greek, Dutch and others, all below 2% respectively.
      Now, the “Australian” category covers people who call themselves “Australian.” If you were to split this, you’d find that they’re mostly British or Irish and have been in the country for multiple generations.
      Another way to look at who is who is to count their religious affiliation, for example, Catholic versus Protestant. Here, Wikipedia tells us that 25.3% of Australians are Catholic, 17.1% Anglican and 18.7% “other Christian”. If you add non-Catholic Christians, you get 35.8%, which is close to the 36.1% of the English share of the population. By the same token, 25.3% – 10.4% = 14.9%. So, these could be the Irish hiding under the label of “Australians”.
      Assuming that (nearly) all those labeled as “Australians” are either British or Irish, we find that the total number of Australians of British and Irish descent is 90.8%. How then can “others” add to nearly 20%? This is because many people will declare multiple identities. Depending on how the question is asked, they may answer that they consider themselves “Australians” and that they have, say, one Greek parent and one English one. They would then be added to all three categories, thus the percentages add to more than 100%! A visible proof that it’s hard to slot people into boxes.
      In any case, Australia is very strongly British/Irish, much more so than Canada and the US. The largest plurality in the US today are Germans, followed closely by Hispanics (Spanish is a de facto second official language in the US today), then African Americans, then only we have the English (with pluralities in New England and Utah–yes, the Mormons are mostly of English descent), the Irish, and others in smaller proportions. There is also a large category in the US calling themselves “Americans.” They concentrate mostly in the Appalachians, and, when you dig deeper, they turn out to be mostly of British (Scottish and Welsh rather than English) and Irish descent, whence their traditional Bluegrass music and accents derive.

  • Peter Kemmis says:

    Judith Curry recently posted this on her own site; she refers to Dyson’s perceptions about facts and theories, which seem highly relevant to Don’s post:

    “With regards to climate science, the biggest concern that I have is the
    insistence on ‘the facts.’ This came up during my recent ‘debate’ with
    Kevin Trenberth. I argued that there are very few facts in all this, and
    that most of what passes for facts in the public debate on climate
    change is: inference from incomplete, inadequate and ambiguous
    observations; climate models that have been demonstrated not to be
    useful for most of the applications that they are used for; and theories
    and hypotheses that are competing with alternative theories and

    I particularly like Dyson’s clarification on facts vs theories:

    Facts and theories are born in different ways and are judged by
    different standards. Facts are supposed to be true or false. They are
    discovered by observers or experimenters. A scientist who claims to have
    discovered a fact that turns out to be wrong is judged harshly.

    Theories have an entirely different status. Since our understanding
    is incomplete, theories are provisional. Theories are tools of
    understanding, and a tool does not need to be precisely true in order to
    be useful. A scientist who invents a theory that turns out to be wrong
    is judged leniently. Mistakes are tolerated, so long as the culprit is
    willing to correct them when nature proves them wrong.

    The loose use of ‘the facts’ in the public discussion of climate
    change (scientists, the media, politicians) is enormously misleading,
    damaging to science, and misleading to policy deliberations.

    I would also like to comment on the ‘good loser’ issue. I
    wholeheartedly agree with Dyson. In the annals of climate science, how
    would you characterize Mann’s defense of the hockey stick? Other good or
    bad losers that you can think of in climate science? The biggest
    problem is premature declaration of ‘winners’ by consensus to suit
    political and policy maker objectives.”

    ref: http://judithcurry.com/2014/04/21/the-case-for-blunders

  • John Morland says:

    Astronomy has had its examples of protective the hypothesis, and no doubt you could find other cases in other scientific disciplines. Certainly this happens when the stakes are high, inflated egos with careers on the line. The debate over the Big Bang v Steady State is an example.

    An earlier example, Sir Isaac Newton wrote the “Principia Mathematica” (published 1687) wherein lies the famous gravitanional formula F=G m1.m2D^2 (where G is the universal gravitanional constant (6.674×10^-11 m^3kg^-1S^-2 – a very small constant which equates gravity being by far the weakest of the 4 natural forces–but I digress). Newton also invented calculus. The upshot was he thought himself the best of his time (and indeed he probably was or, if not, a key figure in the scientific revolution) and acted accordingly. However, he came a bit of cropper when he tried to pin down the Moon’s orbit and its long term motion.

    His calculations to predict the Moon’s position although close, was not accurate enough; there was a positional drift over time. First he blamed the observers – they were inaccurate, wrong even incompentant, then he blamed the equipment (telescopes, mountings etc) and its designers/ engineers. In the end he had to refine his calculations to little avail and the blame cycle continued.

    He never accurately solve the Moon’s orbit. Although his basic gravitanional formula was right, he considered the problem as Earth and Moon gravitanional interaction, when really it is Sun Earth Moon interaction.

    What is different now in AGW debate is the unwillingness to disclose the raw data and methods used to arrive at a CAGW position.

    • David says:


      You can’t have it both ways. Previously you have claimed that your “special friend” at the ABS declared the statistical methods used by the IPCC “did not did not reconcile with established statistical methods”. You wrote

      “Before he [Australian Statistician] retired, the department reviewed the IPCC’s then reports on AGW (possibly AAR 4) Their review showed the IPCC’s data, and the conclusion reached, did not reconcile with established statistical methods. In other words, the ABS’s view was the IPCC’s conclusions and predictions could not be statistically supported based on the data provided.”


      But today after you got out of bed you decided to post

      “What is different now in AGW debate is the unwillingness to disclose the raw data and methods used to arrive at a CAGW position.”

      Either the data and methodology reviewed by the IPCC to support AGW is available for scrutiny or it is not. Which is it?

      • John A Morland says:

        Firstly David, I wish you would disclose you family name, like I do and others do on this blog site, as a matter of courtesy if nothing else.

        Secondly, you know very well that data from official IPCC AAR reports is disclosed but other data from individual CAGW “scientists” are vigorously defended and require court action to be disclosed, by way of example the recent Michael Mann? UVa case (which he won).

        In future David I will not respond to a challenge or query from you on this unless you disclose your full name.

        • Gus says:

          “I wish you would disclose your family name…”
          But why? I disclose my face instead, whereas you don’t. I wish you would disclose your face!
          “On the Internet, nobody knows you’re a dog.” Yes, I am a dog and proud to be one. Show me another dog who can type and argue with humans in English, even if somewhat tainted with Bark. It took me ages to learn how to push the keys with my paws. Doing “Shift-U”, I can tell you, is quite a challenge!

        • David says:


          The reason I don’t post my family name is that I am still in the workforce. I could give you a false name, but that would be silly. I agree with Gus. I can’t believe I just wrote that 🙂 This blog is all about the ideas. If you make a sweeping statement, I am going to comment on it.

      • GenghisCunn says:

        Two things here. 1:
        the IPCC projections for 2100 all depend on econometric modelling. Their modellers were not experts in national accounts statistics, and their work was deeply flawed. The ABS was a pioneer in developing NAS, and the internationally-agreed NAS standards are based on the work of the ABS,
        including work done when Ian Castles was director. (I once gave Castles credit for this, he gave credit to the ABS.) The Castles-Henderson
        (former OECD Chief Economist) critique totally discredited the IPCC’s
        work. Some at the IPCC welcomed this, as assisting them to present valid data. But the hierarchy pulled down the shutters, the work was never corrected. I recall that under one scenario, the GDP of South Africa in 2100 exceeded World GDP in 1990. Enough said.

        2. Many of the scientists whose work forms the basis of IPCC reports have been very reluctant to share their data – Mann is not an outlier, but typical in that respect. Their work depends heavily on statistical work in which they lack expertise, and their data handling is atrocious –
        Climategate revealed that it would be impossible to replicate the work of many of the leading scientists. Most work is funded by governments who require data sharing, but this has been fought tooth and nail – witness long-running FOI battles with CRU.

        Mr Morland, my name is Michael Cunningham. For reasons which I won’t go into, I have for years posted online mainly as Faustino or Genghis Cunn. These names are well-known, like a brand name. The best known Michael Cunningham is a New York novelist, I’ll stick to my noms de net, thank you.

      • GenghisCunn says:

        PS. No, I am not a blonde hedgehog.

    • Gus says:

      “What is different now in AGW debate is the unwillingness to disclose the raw data and methods used to arrive at a CAGW position.”
      Well, this is just Mann, who, frankly, is a freak… but you have to be rather careful saying this: he likes to sue. Nobody takes his hockey stick curve seriously any more, not even his close collaborators. It is widely accepted that the Medieval Warm Period was likely as warm as today or warmer and global, that there was the Little Ice Age, global too, the Roman Warm Period and so on. Every real scientist shares data… on publication. Many leading journals demand it.
      Just as importantly, scientists from countries other than the US, UK and Australia have been contributing to the common pool of knowledge for quite some time now, adding their data, which is open too. And so, you can find much about historic temperatures in Tibet or Manchuria from Chinese scientists, about temperatures in Kamchatka from Russians, about water level in Parana from Argentinians. None of this data agrees with the Mann’s “hockey stick.”
      It really should be understood that the few names like Mann and Hanson do not stand for “climate science,” and are not representative of it or of its practitioners. These people make a lot of noise, but they are propagandists, not scientists, working to advance their ideology, but it has nothing to do with science.

Leave a Reply