I was wanting to briefly highlight a David Roberts post called the two key points that climate “skeptics” miss. I was only going to comment on one of the points, which was about the convergence of all the evidence, typically referred to as consilience. Given the continued fighting about consensus studies, it’s easy to forget the real reason why our understanding of climate change is robust; it’s because of the combination of all the evidence that leads to a coherent, and consistent, picture of the likely consequences of anthropogenic CO2 emissions.
Of course, this doesn’t mean that our understanding is perfect, or that there isn’t uncertainty. The basic idea is essentially that all the different lines of evidence are consistent, even if not precisely the same. Also, not each line of evidence will be equally strong, and there will still be uncertainty associated with each line of evidence. This is where the problem comes in. A typical “skeptic” tactic is to attack, and focus on, a particular line of evidence. Adjustments to the temperature data. The “pause”. Ocean heat content data coverage. The hockey stick. Climate models. Of course, if you focus on something specific, you probably can find something to criticise, even if it isn’t particularly significant. It’s highly unlikely that any single line of evidence will be perfect, and so if you really want to find fault, it’s probably quite easy to do so.
credit : Gavin Scmidt
In truth, my main reason for writing this post was to highlight the graph above and to put it into some kind of context. Having done so, I will now stop.
...and Then There's Physics 
OT, not really, but everybody needs to go read Victor’s Variable Variability post on the fall AGU meeting. It’s rich.
Is consilience a weaker concept than weight of evidence?
ATTP,
I see one graph in your post, but it doesn’t look like the one Schmidt et al. published with their 2014 Nature Geoscience Commentary Reconciling warming trends, which shows how accounting for ENSO, solar activity, and anthropogenic and volcanogenic aerosols brings CMIP5 model projections much closer to GISTEMP and HADCRUT4 observed temperatures. I don’t know how to embed their fig. 1b here, unfortunately.
I think Gavin had the graph on tweeter just recently.
On the contrary, consilience is a much stronger concept.
See for instance https://en.wikipedia.org/wiki/Consilience
Consilience in climate change is remarkable, covering a huge range of different processes, all of which are consistent with the earth heating due to greenhouse gases.
We have:
The basic physics: the radiative properties of CO2 and H20
Simple modelling: 1d atmospheric models coupled with the vapour pressure of water
Surface temperature records showing rises consistent with the physics
Rising sea levels accelerating as expected with warming
The retreat worldwide of mountain glaciers
The collapse in Arctic Ice volume
The cooling of the stratosphere, as expected
The changing of species’ range, expected with rising temperatures
The rise in ocean heat content
It is an astonishing set of consilient evidence.
Weight of evidence would come in, for instance the surface record where quasi independent analyses of the same property (HADCRUT, GISS, NOAA, JMA, BEST) all add weight to the same conclusion.
What’s the weight of evidence, and does it have something to do with Al Gore?
Hmm, in the cited figure of Gavin Schmidt I see a trend slope of the CMIP5 ensemble mean of roundabout 0.25K/decade after 1995. And I see a very similiar slope of 4 undependend datasets of 0.1K/decade in the same interval for the real world tropics. What do you see? That the very low end of the models still match the observations? Which models do this? What is their TCR? The TCR of the CMIP5 ensemble mean is near 1.9…
VTG and Willard
Thank you for your replies. It is some of the text in the wikipedia article that raised the question in my mind:
That is, when multiple sources of evidence are in agreement, the conclusion can be very strong even when none of the individual sources of evidence is very strong on its own.
The bold in the quote is mine and is the phrase lead to think differences if any. This of course leads to Willard’s question, What’s the weight of evidence,…? Well, this first part anyway. I could have easily asks is ‘consilience stronger than weight of evidence?’
Perhaps a better way to ask is “How do consilience and ‘weight of evidence’ differ? VTG, for you is independence a defining difference? Or perhaps weight of evidence was/is too broad of a concept.
In any case it seems to be an interesting question (to me).
MW,
what’s interesting to me is just how wide a range of evidence all supports the same theory.
I forgot to mention ice cores. Amazing what’s been done with paleocliomate, no?
vtg – don’t forget lake ice on’off dates, inland lake water temperatures, and dozens of phenological studies
mwgrant, “weight of evidence” = “balance of probability”. It implies that while different lines of evidence may vary as to whether they indicate a theory is true, ie, with some indicating it is true, and some that it is false, and some indifferent, conjointly they indicate that the probability of the theory being true given the evidence is at least better than 50%.
In contrast, in a consilience of inductions all (or nearly all) lines of evidence must support the theory by at least showing it to be better than chance that it is true; with the conjoint evidence strongly favouring that theory over other proposed theories.
Frank,
I think you’re both missing the point, and illustrating it, at the same time.
Mal,
Gavin’s tweet is below. I think his paper is for surface temperatures. This is intended to be a proper comparison with satellite temperature datasets.
Good point Kevin.
Then there’s the recently directly measured surface forcing due to CO2
http://www.nature.com/nature/journal/vaop/ncurrent/full/nature14240.html
Perhaps we should get an official body to put all this together in one authoritative report, what do you think?
ATTP:
Ah, OK. IMO though, the tweeted graph does appear to support the AGW-denier meme of “the hiatus”, i.e. observed temperatures rising more slowly than model projections. OTOH, Schmidt et al. 2014 makes the point that climate science is hard at work refining the models by resolving short-term “noise” to forcings. Referring back to this RC post:
Statistical analysis has answered “no” to questions 1 and 2, while Schmidt et al.’s Commentary addresses questions 3 (yes, a slight slowing between 1998 and 1913) and 4 (short-term variation forced by ENSO, reduced solar activity, aerosols, etc).
It comes down to, as you say, AGW-deniers’ typically dishonest representation of how Science works. One wishes scientists wouldn’t make it easy for them, though.
In my last comment, there’s a double-quote missing at the beginning of the link to Schmidt et al. 2014.
Mal,
I fixed the link to Schmidt et al. in your earlier comment. I agree that it might support some of the “pause” rhetoric, but I still think it’s a nice graph as it pretty solidly rebuts the Christy figure that is regularly hauled out to supposedly highlight that models have failed.
Tom Curtiss,
Thanks, Tom. This really seems to be a rich topic.
To paraphrase – nothing in Biology makes sense without evolution, nothing in climatology makes sense if CO2 is not a greenhouse gas. (And feedbacks are positive, sensitivity 2-4K/doubling, see rest of AR4 for details but this kind of spoils the snappiness)
I followed here from Gelman’s blog. I am still stuck on this averaging process to get a value for climate sensitivity and would appreciate discussion on this topic (maybe it is even worth it’s own post, I dunno). Say I have maps of forcings F and temperature T at time 1 (F1, T2), and the same at time 2 (F2, T2).
The simplest version of the climate sensitivity calculation is of the form lambda=deltaT/deltaF, where delta indicates the difference. There are no feedbacks, etc. I believe this is sufficient for me to explain the conceptual problem I am having. There are many ways this can be calculated and appear to fit the definition, lambda can equal:
A. [mean(T1)-mean(T2)]/[mean(F1) – mean(F2)]
B. mean(T1 – T2)/[mean(F1) – mean(F2)]
C. [mean(T1)-mean(T2)]/mean(F1 – F2)
D. mean(T1 – T2)/mean(F1 – F2)
E. mean((T1 – T2)/(F1 – F2))
Here T1-T2 means subtract value at coordinate [1,1] in map T2 from the corresponding value (at coordinate [1,1]) in map T1.
Is the same calculation used in all papers? If yes, which one is it and what is the derivation showing it is based in physics?
Anoneuoid,
The standard energy balance formalism is:
where
is the system heat uptake rate, 
is the change in forcing, 
if the feedback parameter, and 
is the change in temperature. All of these are essentially determined by assuming that the planet is a single system. For example, to determine 
you determine how the total energy changes with time, if you introduce some external change (add GHGs, for example). Similarly for 
, which is normally determined by averaging the change in temperature at every location (anomalies).
I don’t actually know how to explain this any more clearly, and we have been through it a number of times on Andrew Gelman’s blog. It is simply an approximation that allows one to do basic estimates and I’m not even entirely sure what your issue with it is.
You may find my earlier comment at Climate Etc. interesting. Hope to make it into a post one day. It fits to the current consilience of evidence post.
Given that the difference [between models and microwave temperature trends] is mainly due to the missing tropic hotspot in the satellite temperature trend, it seems more likely than not that there is some problem with the satellite trends.The tropical hotspot 1) is seen in some radiosonde datasets, 2) it is seen in radiosonde winds, 3) it is expected from basic physics (that we know that the moist adiabatic temperature profile should be a good approximation in the topics due to a lot of convection), 4) you see the strong response of the troposphere compared to the surface at shorter time scales and 5) it is seen in climate models.But we will only know this with confidence when we find the reason for the problem with the satellite trends or when we find problems with all of the other 5 pieces of evidence against it.
JCH says: OT, not really, but everybody needs to go read Victor’s Variable Variability post on the fall AGU meeting. It’s rich.
Thank you JCH. Here is the link:
http://variable-variability.blogspot.com/2015/12/my-theoretical-agu15-program.html
Some tropospheric temperatures, homogenization, citizen science and changes in variability. Looks like AGU will be interesting.
@aTTP, Thanks for providing the forum for this question. I do not think it is as trivial as you are treating it. You can help by simply answering the questions I asked. To simplify I set N=0, and I guess my lambda is the reciprocal of your lambda. That is fine, I am not worried about getting correct values here, only what differences are being taken. It is still not clear to me from your post which calculation is used (or if these vary from paper to paper).
The description using “Total energy changes” and “averaging the change in temperature at every location” is still too vague for me to understand. I am looking for someone to pick an answer from those I provided or write in their own of the same form.
Anoneuoid,
I don’t think that one of yours is necessarily selectable. As far as the energy terms are concerned, they’re straightforward. Treat the entire planet as a single object. Calculate how the energy changes with time, divide by the area. The temperature is somewhat more complex, but my understanding is that the norm is to determine some kind of average for each location. You can then determine the change in temperature for each location. The average for the planet is then the average of the change for every location. Alternatively you could determine the surface flux and then use the blackbody equation to determine the equivalent temperature. The change in temperature would then be determined by the change in surface flux.
” Of course, if you focus on something specific, you probably can find something to criticise, even if it isn’t particularly significant. It’s highly unlikely that any single line of evidence will be perfect, and so if you really want to find fault, it’s probably quite easy to do so.”
If the glove doesnt fit, you must aquit.
The general tactic has been to pile on lots of evidence in the hopes that folks will look at the “weight” of evidence. but, the problem comes up of the glove that don’t fit.
So here you have a glove.. its not particularly strong as a piece of evidence. by itself it would do nothing to further the case.. but you throw it into the mix. and it back fires..
When it backfires folks start to doubt the other evidence… after all you are the expert and why are you putting this stuff forward.
The hockey stick is the glove
Steven Mosher:
Maybe in a court of law.
Is that the best you can do? The jury finds your client guilty as charged.
I believe the appropriate analogy to the hockey stick is:
“If the glove won’t go on the hand when the hand is deliberately being held rigid, with stiff fingers held wide apart, then lawyerly types will say you must acquit.”
(Note: I am not claiming this is what OJ did. I am claiming his lawyer’s arguments were simplistic and based on easily gamed evidence.)
I believe the appropriate response to the hockey stick is:
Which one?
A glove weights nothing compared to a 150 kg Wookie:
Consilience in anthropogenic climate change may not be as strong as some think. Consider VTG’s list of processes from above:
The basic physics: the radiative properties of CO2 and H20
Simple modelling: 1d atmospheric models coupled with the vapour pressure of water
Surface temperature records showing rises consistent with the physics
Rising sea levels accelerating as expected with warming
The retreat worldwide of mountain glaciers
The collapse in Arctic Ice volume
The cooling of the stratosphere, as expected
The changing of species’ range, expected with rising temperatures
The rise in ocean heat content
Parsing out anthropogenic contributions from natural contributions would have to accomplished in each case. Independence in the ‘measurement methodologies’ has to be addressed, demonstrated, and documented–both the original sets and the parsed out anthropogenic sets. Modeling of course should be eliminated if it is not V&V’ed. (IMO it is unlikely that simple 1d models could be.) I agree with “I do not think it is as trivial as you are treating it,” [Anoneuoid] but for very different reasons. Consilience is still interesting because it suggests some a line to maybe get more insights.
My point is that I do not think one can evoke consilience as supporting man-caused climate change until more is accomplished and documented. While interesting this specific evocation probably is not ready for prime time.
‘Prolly shouldn’t say anything here as this discussion is above my pay grade, but I’ve never let that stop me before.
iIt seems to me that consilience is important in that it speaks to risk – in addition to whatever extent it speaks to confidence, such as confidence in a range of sensitivity.
If recent warming were not anomalous, what is the likelihood that so many measures would point in that direction (even if each individual measure has a significant degree of uncertainty). That doesn’t mean that the individual lines of evidence fit together in aggregate to strengthen each other in a point of convergence, and it does’t mean that the existence of other lines of evidence pointing in the same direction strengthens the certainty of any one line of evidence – but that so many lines of evidence pointing in the same direction mean that the probability of their being no risk is lower.
‘Prolly shouldn’t have said anything. 🙂
There’s no shared alternative framework, just a fixed certainty that the consensus must be wrong.
If the mainstream scientific consensus on anthropogenic climate change is wrong, then we’ll need some other theory that makes sense of present-day changes and harmonizes with data from historical record. Climate skeptics have offered no such theory. Where climate change science is fecund, climate skepticism is moribund, merely destructive. …
Natural variation can’t do it.
Joshua, IMO consilience, or even characterizing the consilience in a particular case, certainly ties in with risk. However, it seems that at the formal level it is not without pitfalls or constraints or for that matter, effort. This is based on what little I have read about it to this point. I expect my view to continue to evolve.
Roberts’ Vox post and ATTP’s online descriptions are limited. It would be interesting if there is a more formal application or documentation specific to climate change lying around somewhere.
> Parsing out anthropogenic contributions from natural contributions would have to accomplished in each case.
If we associate positions regarding such cases, we get the opposite of consilience. If we replace positions with actors, we may get a divide and conquer mechanism:
Click to access Vermeule_639.pdf
Tertius Gaudens is a related but lukewarmer strategy.
Willard,
If we associate positions regarding such cases, we get to the opposite to the idea of consilience…
That was and is a concern. On the other hand it would seem that lists as in Roberts graphic and VTG’s comment do not provide unique information/measurement on the anthropogenic component. If consilience is demonstrated (in some QAed formal documented process) one would have provided an argument for global warming* but what does that accomplish in the context of developing policy regarding the man-made component. A formal methodical look at consilience would be interesting.
> If consilience is demonstrated […]
Consilience is not a formal deduction. It’s an evidential inference. Regarding AGW, one simple formulation I know is Richard Alley’s:
Everything we know points toward man-made CO2 as a principal cause of AGW. When you consider all the lines of evidence with this main culprit in mind, everything makes sense. That’s the best explanation we have. In fact, that’s the only explanation that coheres with our evidence basis.
I suppose one can try to recycle the “but V&V,” but I don’t think it will lead you very far. It did not the last time. Why would it now?
***
There’s an interesting aside regarding Richard Alley and hockey sticks:
The consilience between hockey sticks audits and lukewarm playbooks is quite thin.
mwgrant, the AGW theory can be summarized as saying:
1) Burning of fossil fuels will result in an increase in CO2 concentrations in the atmosphere;
2) CO2 is a greenhouse gas;
3) Therefore there will be an increase in Global Mean Surface Temperature.
These claims can be considered to be the “hard core” of the AGW “scientific research program”. That hardcore is fleshed out with models of various depths. Without those models, the hard core makes just two predictions, and not even those two without ceterus paribus clauses. With the models, the range of predictions expands, which means the range of potential phenomena explained by the theory also expands. It is only with the expansion of predictions that we can have a consilience of inductions.
What you are trying to do by “parsing out the contributions” is to split the theory into to parts to prevent any such consilience. So, where AGW predicts:
1) Rising CO2;
2) Rising GMST;
3) Falling diurnal temperature difference;
4) Falling seasonal temperature difference;
5) Falling latitudinal temperature difference (polar amplification);
6) Expansion of the Hadley cells;
7) Rise of the tropopause;
etc
you consider just one prediction (rising GMST) and its corrollaries and say we must parse out the “natural contribution and the anthropogenic contribution”, thereby preventing the consideration of the full range of evidence that AGW consiliates.
Worse, ‘natural’ is just a place holder for ‘not anthropogenic’, so the “core” of NGCC (Natural Global Climate Change) is that:
1) The sun gets brighter; or
2) The sun gets dimmer; or
3) The sun retains its brightness; or
4) Galactic Cosmic Rays modulate cloud albedo positively; or
5) GCR modulate cloud albedo negatively; or
etc, etc, etc.
It is very clear that the theory of NGCC predicts everything and nothing. It is in fact not a theory at all.
The upshot is, while there are reasons to distinguish between anthropogenic effects and effects due to insolation changes, effects due to volcanism, and so on, those distinctions are not relevant to the concilience of inductions. More importantly, if you want to propose a theory that better conciliates the inductions than AGW, you actually have to propose a theory. You can’t use neoscholastic place holders like “natural” as if that somehow represents a theory.
And of course, once you do actually propose a theory, you find that either it does not conciliate many inductions at all – or that you need a complex compound theory where multiple ad hoc hypotheses are joined together to explain what AGW explains fairly straightforwardly.
“The consilience between hockey sticks audits and lukewarm playbooks is quite thin.”
Which is, of course, because climate contrarians typically do not propose an actual theory, and consequently do not impose on them selves the requirement of consistency.
“It’s natural” is not more an explanation than “dormitive power”.
With respect to Gavin Schmidt, the plot shown is underwhelming (and I’d question the vertical scale used), even if Christy played fast and loose with the baselines in his own plot:
A better figure is the surface temperature anomaly (measured & modeled) from Schmidt, Shindell and Tsigaridis (Nature Geoscience, March 2014). I’ve taken the liberty of extending GISTEMP LOTI to 2014 and the most likely 2015 value. It still looks like we’re ‘missing’ about 0.15°C.
Willard,
Yes it makes sense. But is it a convincingly unique ‘solution’ resulting from a formal process? It would be interesting if someone had made an attempt and published.
#####
A couple of reasons for a more formal approach to consilience…
1.) Induction of the lines of evidence is the heart of the approach. Thus it would seem that in a matter of global concern, e.g., good detailed documentation for the induction of each line is necessary. After-all the purpose is to make a solid case.
2.) Clarity and structure. Note what ATTP wrote:
it’s because of the combination of all the evidence that leads to a coherent, and consistent, picture of the likely consequences of anthropogenic CO2 emissions.
The assertion that consilience “leads to a coherent, and consistent, picture of the likely consequences of anthropogenic CO2 emissions.” [my bold] is distinctly different than your assertion “Everything we know points toward man-made CO2 as a principal cause of AGW.” While not a great concern here that does show the ease of loosing clarity and why even qualitative heuristics are better done in a formal documented methodical manner. Formal methodology assists in a a clear formulation of the problem. Compiling a list of processes/observations or constructing a graphic of the the same and then evoking consilience is not a viable, convincing approach.
#####
My reference to V&V was only to the simple 1d models in VTG’s list. Not to other LOE inductions.
There are also two more extended plots from Schmidt that appeared in this Guardian article by John Abraham back in August:
http://www.theguardian.com/environment/climate-consensus-97-per-cent/2015/aug/10/2015-global-temperatures-right-in-line-with-climate-model-predictions
mwgrant – the *consequence* of anthropogenic Co2 emissions *is* global warming. Which is exactly what the second sentence you quoted says. How in the world do you read them as distinctly different?
Willard… I see ducks and ducking…
Tom Curtiss and Willard
Tom
I read your 12-14 3:12AM summary before my last comment. It is exceedingly helpful. It is obvious–we turn the dial on the CO2 [Willard’s video …Doh!] and [detrimental] parsing is not necessary.
Willard
You hit the mark with the videos and comments.
Tom and Willard — really,really good stuff today. A lot of turns. And still things to think about. Thanks!
mwgrant:
“It would be interesting if someone had made an attempt and published.”
Click to access RoyeretalCO2GSAToday’04PhanerozoicClimate.pdf
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.372.1389&rep=rep1&type=pdf
Richard Alley’s talk is based on the second of these papers, with some updates showing some of the apparent discrepancies disappearing or becoming dubious with finer resolution stratigraphy.
==> “When it backfires folks start to doubt the other evidence…The hockey stick is the glove”
There were lots o’ folks looking to invalidate the evidence way before the hockey stick. And their “motivation” to find evidence that confirmed their bias was not because they saw evidence that didn’t fit.
The proof is that they accept, without due skeptical diligence, other evidence that doesn’t fit. They don’t particularly care about the fit of the evidence.
Magma
No. Willard writes the case to be made via consilience is CO2 emissions are the driver. ATTP writes that the case to be made is the consequences of the CO2 emissions.
> ATTP writes that the case to be made is the consequences of the CO2 emissions.
I think what AT says coheres with what Richard Alley says because it follows from it. Here’s another Richard Alley presentation on the likely consequences of AGW:
Source: http://whatweknow.aaas.org/richard-alley/
***
Here’s something I think we have to bear in mind when we want a more formal process:
http://www.britac.ac.uk/events/2007/evidence/abstracts/tillers-long.cfm
In other words, scientific explanation is still a creative act. (The rest of the abstract looks like convoluted and I suspect incorrect platitudes, but that part sounds right.) The law could be greatly improved if we had evidential inference engines. Just imagine if law inferences could be processed through free packages like we find in R. The first one I’d like to see is the Supreme Court Deluxe.
Informal reasoning works in an open world, where you seek to satisfice from what you got. Formal reasoning oftentimes assumes a closed world, if only because it helps connects the dots in a tractable manner. In between, there are ways to formalize without really trying to automate anything, like MCDA:
https://en.wikipedia.org/wiki/Evidential_reasoning_approach
All the different lines of evidence are in the IPCC reports.
Let’s make it easy and give every line of research the same probability. And lets be pessimistic for most and only give the 90% chance of being “right”. Then one line of evidence has 10% chance of being wrong, 2 lines have 1% chance of being wrong, 3 lines, 0.1%, and so on. That is the statistical side, this can be done formally.
For me the consilience of the evidence is mainly a protection against unknown unknowns. I would thus have a strong preference for 3 lines of evidence with 90% chance of being right over 1 line of evidence with 99.9% of being right. The latter could be toppled by finding one unknown unknown. That there are 3 (or in our case a lot more) unknown unknowns of the size that would topple a line of evidence is very unlikely. Especially after decades of research, by many different groups in a mature science.
Because unknown unknowns are unknown, the consilience of evidence is fundamentally subjective, but still a very powerful concept, especially when researching a complicated complex system having mainly observations and only experiments for partial aspects.
mwgrant,
I don’t know what you’re getting at. All I mean by this is what will probably happen, given our understanding (warming, sea level rise, reductions in land and sea ice, changes to the water cycle, …). I don’t mean anything more than that.
Mal,
I agree that the surface figure is good, I just liked that Gavin had produced a figure that addresses the Christy one that gets dragged out over and over again. It’s also interesting to see who (on Climate Etc) is getting upset that I’ve highlighted this.
The thing that strikes me about the Gavin’s tweet is that even if the trends were the same, the variation in the observations would only just lie within the spread of the models. Unless the observations we have are a particularly wiggly realization of the climate, that suggests to me that the models may be underestimating the effects of natural variability somewhat, although that is just my intuition.
Apart from the fact that the y-axis could be scaled better, it looks like a good diagram to me. I think it is worth noting that good scientists like Schmidt have no problem discussing data that suggests an issue with the models, where that is what the data show, which is a marked contrast to those who only want to discuss the evidence that suits their position (and even then provide a highly “nuanced” presentation, that will mislead the general public).
MW,
I’m not sure why you think natural factors should be warming rather than cooling.
To pick out specific independent lines of evidence pointing to the anthropogenic nature of the change is, of course, possible:
for instance:
OHC tells us the imbalance is radiative rather than from internal rearrangement
Insolation has not changed significantly over the period
Forcing from CO2 has been directly measured
The spacial pattern of warming is consistent with CO2
The magnitude (>1c) and rate of change are high compared to paleoclimate changes (glacial/interglacial)
Basic physics of the greenhouse effect
Stratospheric cooling
etc
It’s interesting to note your “VTG’s comment do not provide unique information/measurement on the anthropogenic component.”
Your demand for “uniqueness” of measurement cannot possibly be met in a system with multiple drivers, even in principle. It’s a great example of setting impossible demands, common in the climate blogosphere.
BTW when I say “issue with the models”, I don’t mean the models are necessarily wrong, just that there is an apparent discrepancy between the models and the observations, but that doesn’t mean the problem isn’t with the observations (or a bit of both).
In a sense, this is the point. The whole point of consilience is that a vast majority of the evidence points to the same basic conclusion. That each line might have weakness doesn’t change this. Focusing on potential weaknesses, rather than considering the overall position, is the point I was getting at in the post. Also, that one can think of some alternative, doesn’t suddenly mean that that alternative is somehow plausible. You really need evidence to support that alternative, not simply point out that it hasn’t been entirely ruled out.
Dikran,
I suspect that Gavin simply chose a y-axis range that was symmetric about 0. Would probably have been better to go from -1 to 2, rather than -2 to 2.
ATTP, indeed, it could be better, but it isn’t a really serious problem ;o)
That is indeed a classic.
If I may add a little bit…
William Whewell, the fellow who coined the term “scientist” in 1833, also coined the phrase “consilience of inductions”.
There seems to be a vagueness in some of this discussion about whether the various elements which come together to form a consilient constellation are all of roughly equal importance -kind of an implied empiricism. One of Whewell’s arguments was that there is a definite hierarchy of information, and that even a consilience of empirical data is insufficient to build reliable knowledge without another concept he introduced: “colligation”.
“Colligation, according to Whewell, is the mental operation of bringing together a number of empirical facts by “superinducing” upon them a conception which unites the facts and renders them capable of being expressed by a general law.” (from the Stanford Enc) My understanding of colligation is essentially the function theory fulfills, of linking together empirical data by positing causal mechanisms which are not directly observable.
This colligating step is crucial to making a consilience of evidence really meaningful.
First, it imposes a heirarchy of evidence. For example, individual hot years are no more proof of the theory of AGW than of AGW denial -they are contingent from the standpoint of AGW theory. But patterns over longer timeframes are given greater significance by the colligating theory.
Second, the colligating theory will posit the existence of non-manifest causal mechanisms -that is, something we have not yet directly observed. Science would be incoherent if it wasn’t organised in the form of empirical data in orbit around a non-empirical core of theory.
This opens the way to new experimental directions, and again contributes to the hierarchy of evidence, in that it defines in advance what kind of consilient data is decisive, and what is more likely to be incidental.
Whewell kicked these ideas around in the 1840s during breakfast meetings with the likes of Michael Farraday, Charles Babbage, John Herschel, J.S. Mill, Samuel Coleridge and others, and yet his ideas are clearly still really valuable. There is no doubt that deductive methods are necessary for science, but these are only really applicable to closed experiments; all communally shared scientific knowledge must be passed on as a constellation of inductions towards the greatest consilience.
Mark,
Thanks. What you say about induction is interesting. I don’t know if you say the whole debate about the clarity of meaning in the IPCC press conference. One of the arguments that the authors of the paper to which responded made was that they were being inductive, while the physical sciences is normally deductive. My own personal experience is that this is simplistic at best, if not completely wrong. A great deal of the physical sciences is now inductive, rather than deductive, and I wondered if you’d had a chance to think about this.
I agree that it is a dodgy notion to suggest that the physical sciences are normally deductive.
I don’t think there is any way to learn a scientific discipline deductively. Any body of knowledge which is shared by a community over space and time must become a network of inductions, not least because at some point, no individual can know everything, but has to trust someone else’s word (even if that is just a textbook).
I think it is important to distinguish between closed and open systems in the world. When we are able to isolate causal mechanisms, only then can we reason about them deductively. The physical sciences are at least capable of sometimes isolating causal mechanisms via experiment, whereas the social sciences can never do that -but I think all this means is that the physical sciences will tend to include some strong inductions, plus some isolated instances of valid deductions. These still have to come together in the form of core theories, though, and correct me if I’m wrong, but I can’t think of any core theory which one could really call deductive.
But it is extremely rare that any causal mechanism exists in isolation in nature, which means we will always have some noise as well as signal. I personally think the differences between what some folks call the ‘hard’ sciences and the ‘soft’ is really to do with how many opportunities the respective disciplines have to conduct controlled experiments. We then get disciplines with very tight, stable constellations of knowledge -very strong inductions- or disciplines with looser constellations. But I think Whewell was right all those years ago, that we will never get a discipline without a consilience of inductions.
@ mwgrant December 14, 2015 at 5:18 am
I don’t follow your point. My first post was delayed in moderation so its follow-up posted a few minutes later lacked context. Maybe it was misinterpreted.
Magma — that should have been at Kevin O’Neill. Thanks.
“The whole point of consilience is that a vast majority of the evidence points to the same basic conclusion. That each line might have weakness doesn’t change this.”
See: the theory of evolution (with “On the Origin of Species” the prime example).
Willard,
I ‘formal’ I do not mean formal logic but that it handled in the ordered documented manner of the regulatory agencies such as the US Nuclear Regulatory Commission and the US Environmental Protection Agency.
VTG
I’m not sure why you think natural factors should be warming rather than cooling.
Where that thought is conveyed puzzles me. If it is from a list then that list is a cut and paste of your list. If it is use of a term such ‘global warming’, fine. Think climate change is that works.
t’s interesting to note your “VTG’s comment do not provide unique information/measurement on the anthropogenic component.”
Your demand for “uniqueness” of measurement cannot possibly be met in a system with multiple drivers, even in principle. It’s a great example of setting impossible demands, common in the climate blogosphere.
At the time of the comment local solutions were a concern I had. I have gotten past that–missed the obvious. Tom Curtiss and Willard were quite helpful in that regard.
> I ‘formal’ I do not mean formal logic but that it handled in the ordered documented manner of the regulatory agencies such as the US Nuclear Regulatory Commission and the US Environmental Protection Agency.
I’m of two minds on this formalization, mwg.
On the first one, it’s how institutions work, and institutions may be the most important factor in prosperity:
Interestingly, another important factor in a country’s prosperity is geography. Good weather patterns are good for a country, and most if not all poor countries are situated in tropical zones. AGW’s main threat is an economic tropicalization of the world, and I have my doubts that what we need is more formal ways to state the obvious. Actuaries thrive on formal crunching, however, and the first step will be for reinsurers to design an AGW market.
Planning is good, but doing is better.
***
On the second one, there’s an important trap behind bureaucratizing knowledge, a trap that is even more important in private corporations, most of which promote what David Graeber calls an utopia of rules:
http://www.theguardian.com/books/2015/mar/21/books-interview-david-graeber-the-utopia-of-rules
One only needs to deal with a private insurance company once to see the deviousness of the overall populist memes against public institutions, a deviousness the CAGW meme obviously shares. The proposition of new rules is always fun to those who like to play, but in the end those who can game the system win and those who can’t have to struggle with more and stricter rules. Stricter rules that are promoted as leading to freeer and freeer societies, it goes without saying.
Just look at how new ways to organize things power the culture of tech companies. If that’s not enough, think of electoral promises as new game proposals. We’re being suckered in to play the latest games all the time.
***
The memes against institutions and for CAGW seem to be correlated, which hints at the fact that the concept of concilience of evidence goes beyond purely scientific matters.
Willard,
I’m of two minds on this formalization, mgw.
I am too. Do you think it is easy to make the suggestion? I even retired earlier than I intended in part because of the beast.
Well for the record there is stuff in between. Groundwater modeling has quite a bit in common with climate modeling in regard to difficulties and uncertainty but in has an accumulated history of modeling in a contentious public environment. There are many academics who do well consulting though I am sure they take war stories back to campus.
Maybe it is like eating…moderation is advised.
Fair enough, mwg. When AGW will be taken seriously, something what you suggest will happen. Munich Re already started counting. The IMF too:
https://www.washingtonpost.com/news/wonk/wp/2013/03/27/imf-want-to-fight-climate-change-get-rid-of-1-9-trillion-in-energy-subsidies/
So I’ve added the sentence “Actuaries thrive on formal crunching, however, and the first step will be for reinsurers to design an AGW market” to the first part of my previous comment.
aTTP( https://andthentheresphysics.wordpress.com/2015/12/13/consilience-of-the-evidence/#comment-69281 ):
mwgrant,
ATTP writes that the case to be made is the consequences of the CO2 emissions.
I don’t know what you’re getting at. …
Earlier at https://andthentheresphysics.wordpress.com/2015/12/13/consilience-of-the-evidence/#comment-69269 I had written:
*begin quote*
2.) Clarity and structure. Note what ATTP wrote:
it’s because of the combination of all the evidence that leads to a coherent, and consistent, picture of the likely consequences of anthropogenic CO2 emissions.
The assertion that consilience “leads to a coherent, and consistent, picture of the likely consequences of anthropogenic CO2 emissions.” [my bold] is distinctly different than your assertion “Everything we know points toward man-made CO2 as a principal cause of AGW.” While not a great concern here that does show the ease of loosing clarity and why even qualitative heuristics are better done in a formal documented methodical manner.
*end quote*
HTH
##########################
aTTP ( https://andthentheresphysics.wordpress.com/2015/12/13/consilience-of-the-evidence/#comment-69286 )
Your demand for “uniqueness” of measurement cannot possibly be met…
In a sense, this is the point. …
Well your quote is from VTG. I assume that the comment of mine that he was responding to was [ https://andthentheresphysics.wordpress.com/2015/12/13/consilience-of-the-evidence/#comment-69269 ]:
Yes it makes sense. But is it a convincingly unique ‘solution’ resulting from a formal process? It would be interesting if someone had made an attempt and published.
That was not intended as a demand, hence the qualifier ‘convincingly’. As a general rule I try to make observations. make suggestions, ask questions and throw up trial balloons and try not to make demands.
I am on board with the rest of your response.
VTG [mwg correction]
At https://andthentheresphysics.wordpress.com/2015/12/13/consilience-of-the-evidence/#comment-69299 I wrote
At the time of the comment local solutions were a concern I had. I have gotten past that–missed the obvious. Tom Curtiss and Willard were quite helpful in that regard.
The first sentence was on and other matter and should have been deleted leaving:
I have gotten past that–missed the obvious. Tom Curtiss and Willard were quite helpful in that regard.
Again I have gotten past the ‘parse question’. Thanks for your responses and sorry for any confusion.
Victor Venema
For me the consilience of the evidence is mainly a protection against unknown unknowns.
I like that perspective.
Because unknown unknowns are unknown, the consilience of evidence is fundamentally subjective, but still a very powerful concept, especially when researching a complicated complex system having mainly observations and only experiments for partial aspects.
I do not think that subjectivity is a problem. Use of subjective judgements are everywhere. It just needs to be handled and documented properly. (Thinking from the perspective of decision science.)
willard – is the mgw for mwg intentional?
No, J. Will correct.
[No need to lay down the gloves. – W]
New gloves with no latex gloves underneath that have a coefficient of friction and bunch up under compressive forcing
verses old wrinkly dried out gloves that do have latex gloves underneath that have a coefficient of friction and bunch up under compressive forcing
I saw that on the TeeVee live while it was happening and I’m going OMFG the prosecution has no brains (certainly nary a scientist or materials engineer).
[Idem. – W]
Ad homs are not consilient with your argument, Ethan. Please be more consilient.
Willard you can redact the 1st and last sentences, to make it more consilient.
But the very bad glove analogy still stands AFAIK (one piece of evidence, whether one adjudicates that single piece of evidence to be ‘good’ or ‘bad’ which is by definition not consilient, should not take down the entirety of the evidence in a scientific context, all bets are off when slick lawyers go up against dumb as rocks lawyers).
Thanks, Ethan.
One of Moshpit’s points may have been to remind that consilience was a judgment call, and I think we can all agree on this.
Victor Venema December 14, 2015 at 6:59 am
I agree only partly, to my understanding the “unknown unknowns” are part of the structural uncertainty which is part of the assessment of each individual line of research and the models used there. There is no structural uncertainty in combining evidence, in principle just a formal operation, but you have to judge the dependence between them. If you correct for structural uncertainty afterwards you are double counting the uncertainties.
Seems that mwg’s less consilient elsewhere:
http://judithcurry.com/2015/12/13/a-closer-look-at-scenario-rcp8-5/#comment-750742
Willard ,
Track the history and you will seeing my comments evolved at two sites after…and an end is not yet reached. BTW time ordering and context are good things when evaluating multiple comments. [Deleatur – TMBTC]
However, I am comfortable if your bring the matter up. I explore things and enjoy it. I enjoy being in the chase. I have found the exchanges on consilience quite interesting and informative, and I am comfortable with the view that there is a lot left to be said about the general topic of consilience in science.
> BTW time ordering and context are good things when evaluating multiple comments.
Words of wisdom, mwg, words of wisdom.
Time ordering does not contradict my observation, and context only reinforces it.
God speed.
Willard,
Yes, your observation stands. I provided what is behind the my ‘less consilient’ status. If you will, please delete the ‘Live with it’.
Regards,
mwg aka mgw, M-dub
> I provided what is behind the my ‘less consilient’ status.
Indeed you did, specifically on December 13, 2015 at 7:53 pm (Georgia Time):
http://judithcurry.com/2015/12/13/a-closer-look-at-scenario-rcp8-5/#comment-750871
This status could be detected on December 14, 2015 at 2:00 am (UK Time):
https://andthentheresphysics.wordpress.com/2015/12/13/consilience-of-the-evidence/#comment-69263
Formal methodical look at consilience is more than interesting: it seems to be required to be “ready for prime time.”
We both agree that this requirement is far from being ready for prime time too.
I’ve edited your post according to your wishes.
I wonder who would dream up the ‘formal documented process’ required by mwgrant. With a project of this size I’m sure it will only take a few years…
What was it that Willard said before… Ah yes;
https://andthentheresphysics.wordpress.com/2015/10/31/greenland-and-climate-science-denial/#comment-65871
General Interference with Organizations and Production
(a) Organizations and Conferences
(1) Insist on doing everything through “channels.” Never permit short-cuts to be taken in order to expedite decisions.
(2) Make “speeches,” Talk as frequently as possible and at great length., Illustrate your points by long anecdotes and accounts of personal experiences. Never hesitate to make a few appropriate “patriotic”-comments,
(3) When possible, refer all matters to committees, for “further study and consideration.” Attempt to make the committees as large as possible – never less than five.
(4) Bring up irrelevant issues as frequently as possible.
(5) Haggle over precise wordings of communications, minutes, resolutions.
(6) Refer back to matters decided upon at the last meeting and attempt to re-open the question of the advisability of that decision,
(7) Advocate “caution.” Be “reasonable” and urge your fellow-conferees to be “reasonable” and avoid haste which might result in embarrassments or difficulties later on.
(8) Be worried about the propriety of any decision – raise the question of whether such action as is contemplated lies within the jurisdiction of the group or whether it might conflict with the policy of some higher echelon.
Click to access CleanedUOSSSimpleSabotage_sm.pdf
I’m not sure why I’m reminded of that. Not sure at all. 🙂
Willard,
Thanks you for the edit. And yes
I do not think one can evoke consilience [effectively] as supporting man-caused climate change until more is accomplished and documented.
I stand by that position–adding the one word as a clarification. No-one in these comments here or in the referenced articles has stated matters in a way that would change my mind on how effectively the consilience has been implemented. I would add that 1.) I think that can be done and 2.) my knowledge on the approach as a decision tool is evolving and hence, is limited. I am considering it from the perspective that as a decision tool and I am interested in both its strengths and weaknesses. Frankly I think that anyone who uses it or evokes it also should be informed on those matters, and at the time when consilience is evoked it necessary at the same time to lay out the ‘pros and cons’ of the heuristic.
Regarding your second quote
a.) If consilience is demonstrated (in some QAed formal documented process) one would have provided an argument for global warming* but what does that accomplish in the context of developing policy regarding the man-made component.
After a little subsequent back and forth at both sites I arrived at the point:
[BEGIN]
Parsing out anthropogenic contributions from natural contributions would have to accomplished in each case.
No. This is wrong.
[END]
2.) on a formal approach:
Willard wrote:We both agree that this requirement is far from being ready for prime time too.
mwgrant wrote: Maybe it is like eating…moderation is advised.
anoilman,
Maybe you should either not read my comments. You seem to have bad dreams.
I find it interesting that I can have extended reasonable exchanges with a number of folks in both camps. You are not one. I wonder where the problem is.
I for one have no idea what and how you create a “QA’d formal documented process” in science in less than a decade or two, who is supposed to pay for it, and who runs it.
(And yes, I have worked in various places including labs and a bit of pharma)
That’s probably why Oilman finds you annoying, because you’re basically saying “Hey, I’ll believe what you are saying if it meets my arbitrary criteria which haven’t been used in science and climatology and have no way of being met unless someone gives me 5 billion pounds to create it.”
What is wrong with the current setup?
> No-one in these comments here or in the referenced articles has stated matters in a way that would change my mind on how effectively the consilience has been implemented.
I think we could generalize that comment to a ClimateBall ™ law – nothing can be stated in a way to change the mind of anyone.
The notion of implementation hints at an exchange we had in August 2015. Here’s something that I think applies to the “prime time” claim:
https://andthentheresphysics.wordpress.com/2015/08/05/its-more-difficult-with-physical-models/#comment-60517
Four months ago, an argument similar to the one against the “prime time” (so to speak) of teh modulz is being made. This kind of “auditing” request is consilient with the Editor’s request that we clarify “Business as usual,” BTW, e.g.:
http://judithcurry.com/2015/12/13/a-closer-look-at-scenario-rcp8-5/#comment-751251
The Editor’s focus was more on the CAGW meme, however.
Guthrie,
I for one have no idea what and how you create a “QA’d formal documented process” in science in less than a decade or two, who is supposed to pay for it, and who runs it.
(And yes, I have worked in various places including labs and a bit of pharma)
OK you’ve not seen it. That just means your experience is in that particular respect is limited. I worked in nuclear and hazardous risk assessment for several decades where is prescribed and routine.
That’s probably why Oilman finds you annoying, because you’re basically saying “Hey, I’ll believe what you are saying if it meets my arbitrary criteria which haven’t been used in science and climatology and have no way of being met unless someone gives me 5 billion pounds to create it.”
No, based on the comment I have seen and what I have managed to read I simply have suspended judgement on the consilience in this specific case and my views are still evolving–independent as much as possible of what I may ‘believe’. I am not concerned on way on another whether that suspension resolves. Anyway how can you cost something you have not any idea about? You bucket has a hole it. :O)
As far as my beliefs…beliefs usually are not very interesting or useful.
> Anyway how can you cost something you have not any idea about?
I have no idea how nuclear reactors are built, but I bet making formal specifications to insure the safety of their softwares cost more than gitting something from the Interwebz.
Challenge met.
Challenge not met. Also definite out of context. Got serious stuff. Will return.
guthrie says:
December 15, 2015 at 8:02 pm
….
What is wrong with the current setup?
I think we should defer that to a different committee and not be too hasty about making decisions. We wouldn’t want to upset our governments after all. Besides I’m not sure that kind of decision is within our jurisdiction. Perhaps we should adjoin for a month and think about this?
Actually Guthrie, processes go through many revisions trying to get things right. Often ignored is how often they got it wrong, and how much that cost. What’s also overlooked is how hit or miss their application really is. Some companies can do it, and some can’t, yet on paper they both appear compliant. Tossing a new process into the mix is a sure fire way to damage a schedule and there’s no guarantee about how successful it will be. Especially if no one knows whats wrong or what needs to be fixed.
Lastly I have no idea how this applies to research. Often project planning and quality control has something to do with development which is pretty much a known. There’s a reason businesses try to remove the R from R&D.
(I’m only familiar with one formal standard, DOD-STD-2167a. We never fully used it because it was pretty crappy and a waste of time. You can see my work in Iron Man 1… its on the dash of the Hummer at the beginning of the movie.)
> Challenge not met. Also definite out of context.
Don’t go all in, mwg. You may lose that one.
M-dub –
==> “As far as my beliefs…beliefs usually are not very interesting or useful.”
Which of the comments that you’ve posted here have been written independent of your beliefs?
Don’t go all in, mwg. You may lose that one
Don’t worry. I am comfortable, Willard.
Which of the comments that you’ve posted here have been written independent of your beliefs?
Probably somewhere between none and all, J-person.
I didn’t ask how many, I asked which. I’m not much on advanced math, but I can count.
Which of the comments that you’ve posted here have been written independent of your beliefs?
I am unable to answer that. [No math involved. :O)] Why is that a concern or interest?
IMO, one problematic pattern in blogospheric discourse is where people burden others with rules of exchange that are unreasonable, or at least aren’t rules that they apply to themselves.
I think we’re all, essentially, expressing beliefs here with some basic shared assumption that there’s something interesting, or at least compelling, about doing so.
It seemed to me that with…““As far as my beliefs…beliefs usually are not very interesting or useful.” you were implying that others were expressing beliefs (subjective) whereas you weren’t (objective). If I’m right about that, then I think you aren’t being fully accountable for the degree to which your comments, like those of everyone else, are largely informed by beliefs.
So as a challenge to how I interpreted you comment, I’m asking you to point out which of your comments were independent of your beliefs Maybe I should change it to, could you point to a couple of your comments that were independent of your beliefs?
Joshua.
Then given your position and mine we are done. I simply am not interested in beliefs–yours or mine. I am interested in that which precedes belief. Beliefs have baggage. Why would anyone want to ‘argue’ against a belief? Beliefs are personal and likely unyielding.
…[Y]ou were implying that others were expressing beliefs (subjective) whereas you weren’t (objective).
[V]iews … independent as much as possible of what I may ‘believe’ followed shortly by [a]s far as my beliefs…beliefs usually are not very interesting or useful implies nothing about others and how they manage/control their beliefs in sorting things out or making statements. You read way too much into things. I’ll let you cleanup after any silliness that ensues if people chase that comment.
==> “implies nothing about others and how they manage/control their beliefs in sorting things out or making statements. You read way too much into things.”
Well, OK. I apologize for my errors.
==> “Why would anyone want to ‘argue’ against a belief? Beliefs are personal and likely unyielding.”
I have exactly the same question. I don’t fully understand why people do that (I have some guesses)…but it happens in these threads constantly.
And, I presume that you meant it when you said that you’re done…but if you ever reconsider…
==> “I am interested in that which precedes belief”
I find it hard to believe that you’re written any comments here that preceded belief…(if you could point me to any specific ones that did, that would help)…which then raises a question for me: Why would you write comments if they’re not interesting or useful?
Joshua,
We can come back to this later (days). The big driver here for me is a real time crunch.
Quick interim thoughts however:
I find it hard to believe that you’re written any comments here that preceded belief
I agree. All we can do as try and that is why I said “as much as possible” in one of the phrases. Making it worse we can not very well perceive or measure bias in our own statements. (BTW ‘all’ in the original answer ‘between none and all’ would seem to cover that case, your position here.) Not recognizing this is a well-worn cliche flaw of academics, but I consider it universal. Deep down things are co-mingled and in a different way for each of us. [That is a belief. :o)]
To me it is entirely possible to accept, for example, consilience on AGW as a basis for ‘rational’ action(s) and not believe AGW. Belief or maybe more precisely, what one perceive’s as one’s belief, is not necessary for action.
It’s late.
All Gavin did with that graph was stretch the x axis and inflate the recent grey error to include the observations. Sad that he would believe the model error margin is increasing.
Consilience is merely conciliatory unless the field is expanded to include the quantum aspects of CO2 saturation; the evidence that each of the last four interglacials, the hollowscene maximum, and likely the MWP were warmer than it is now; the lack of any congruence besides temperature dependency between CO2 and temperature on any timescale from the Phanerozoic to the current crawl.
By all means, let’s look at the big picture.
gymnosperm,
I’m not sure what graph you’re talking about and I think your idea of what is included and what isn’t is horribly ill-informed.
The last post here, by one “gymnosperm”, reads as a hilarious jumble of pseudo-scientific balderdash (“quantum aspects of CO2 saturation”), irrelevant argumentation based on a convenient acceptance of scientific evidence from ice cores (that past interglacials may have reached higher maxima than the current does not disprove human causation of recent changes), ignorance about the literature on the MWP, wilful mis-spellings (“hollowscene”), and sheer obfuscation and obscurity. Even the nom-de-plume has a certain joke value.
Reminds Eli of “Engineering Level Report” until Bob Grumbine got ahold of it and turned the auditor inside out. Killed that one off
http://climateaudit.org/2011/04/17/the-smug-loop/#comment-264199
TL;DR –
More generally:
http://neverendingaudit.tumblr.com/tagged/engineerilyderiving
That’s been my experience… Equally interesting is that they want to terraform the planet without a detailed plan and instead complain that its not going as planned. If you ask them what the concern is.. they are very evasive. “Something’s wrong!” *hands in the air* Then proceed to offer up the minutia du jour as evidence.