**Note since posting Richard has commented to say the quote (which I couldn’t find myself) referred to something else. So apologies Richard and thanks, because it was about time I wrote an article on this. So even if it doesn’t apply in your case, the general drift is stil valid. Richard:

My quote was specifically (and only) in reference to the claim that the historical temperature records are unreliable.

---

I’ve frequently heard an argument from academics who believe in global warming (Like Rob Wilson of St. Andrews) which runs like this:

Skeptic:	“the models don’t work”
Academic:	“well how do you explain the warming if it’s not CO2?”
Skeptic:	“I can’t explain the warming except by natural variation”
Academic:	“well as you don’t have any explanation, why don’t you just agree with us that it must be CO2”

And today I again read a very similar sentiment from Richard Mallett**:

”If we just say ‘we don’t know what happened / is happening’ then we just give up”

I’ve constantly tried to explain why this is just wrong as a way to view the problem, but I’m not sure I’ve ever probably explained my reasoning in detail. So here it is.

As a physicist I understand what they are saying – “we should be able to explain it”. This is the culture of science: “all things can be explained and it is wrong to give up trying to explain them”.
As an engineer I also know that in the real world, systems are far too complex to be able to explain in detail. So the culture of engineering is:

“if it ain’t broke, don’t fix it”.

To translate that Anglo-Saxon concept into the Latinised/Hellenised verbiage of academia:

when their are no symptoms suggestive of a problem or that one will develop in a system, time and resources are better directed at investigating systems where there are known problems which need attention.

However, the sharp eyed of you will note that I’ve subtly changed the focus. The original statement from Richard Mallett** refers to the internal workings of a system. The engineer concept views a system from outside: “if there are no external symptoms”.
And this is an important difference in the way engineers and scientists view systems. To better understand what I mean, I will bring in an example from those human engineers we call “doctors”. Indeed there are also human scientists, but these are not doctors, but are instead biomedical researchers.
So how do each of these approach the problem of illness in humans?
If a human dies from an illness and someone wants to know why. Do they:

1. Sit them down in a chair and ask them how they are feeling?
2. Or .. cut open the human, start examining the entrails, divine the signs and proclaim their cause of death?

Obviously I’m vastly simplifying the situation, but if a human scientist wants to understand a disease, their first instinct is to “get at” the problem and so they are compelled to cut open the patient and try to get a biopsy of the affected tissue. But if a doctor wants to know what is wrong with a patient, the last thing they do is to physically open them up.
The reason is simple. The doctor/engineer knows they are dealing with hugely complex systems which it is neither worth the time or effort to investigate with the costly procedure of “opening them up and tearing out the innards” and they know that such a procedure is more than likely going to make the system/human worse.
So, let’s re-examine Richard Mallet’s assertion:
Richard Mallett** says:

”If we just say ‘we don’t know what happened / is happening’ then we just give up”

How many times have we been to the doctor and heard them say: “I’m not sure what has caused … X, I’d don’t think it is serious, but come back if it gets worse or if it’s not gone away in a couple of weeks?” Has the doctor “given up”, or have they made a clinical judgement that the best cause of action is to monitor the condition and assess it if it persists or gets worse?
Taking to the extreme, what Richard Mallett** is actually proposing is that each and every time we get an unexplained rash, the doctor should send us to hospital for a full series of tests – and if those don’t work, the philosophy of “we have to know what is happening” then requires biopsies, surgery. Finally the ultimate manifestation of this “we have to know what is happening“, is that we try to induce the condition in other people so as to reproduce the symptoms, because “we have to know what is happening” becomes more important even than morality.
Obviously I don’t believer Richard** is actually proposing that we go to these extremes. I am merely making the point that even he will accept that there are many cases where we have to accept we don’t know “what happened / is happening“
And this is why we have doctors and engineers trained in a whole raft of techniques that don’t require opening up people or dissecting enginers, or shutting motorways for detailed tests because they are used to working in situation where they do not know in detail “what happened / is happening“.

The fallacy of excluding the possibility of the unknown.

But because academics work in a culture where they consider themselves to be the world experts in everything, it is almost a sin to admit they “don’t know”. To admit a lack of knowledge is to “give up” as Richard** puts it. To admit a lack of knowledge for an engineer is a statement which helps decide whether resources need to be put into “knowing” or whether decisions can be made without the knowledge.
But this difference in culture has profound repercussions.
Let us create a simple system:

Y = A + B

After investigation, the scale of them is found to be:

Y = 4 + 2

After more investigation it is determined that A varies as a function of T. So:

Y = f(t) + ?

There are now two ways to model this. There is the engineering way that says that even though we don’t know what B is, that we know its scale and therefore we should include a value for this quantity.
Then there is the academic way, that says: we must know what happened / is happening. And because they must know (and find it difficult to admit they don’t know), therefore there should not be anything they do not know. Therefore what they know is all there is (or more accurately all they will admit). So they will write the equation as follows:

y ∝ f(x) + noise
y ∝ f(x) + noise
y ∝ f(x) + noise
y =α f(x)

Where α is a term meaning this is “noise” – something which is just an inconvenience and which in the academic mind is an irritating problem with the real world and which can largely ignored when it is “averaged out”.
So, what happens we we put back the real world values of “y = 4 + 2” into these two equations?

Engineer:	y = f(t) + 2
Academic:	y = (6/4) × f(t)

Here any academic will be furious and be saying: “not y = f(t) clearly the 2 is “noise” and is just a short term fluctuation which DOES NOT INCREASE the value.
However, I assure you this is what really happens in practice. Here is how it does happen. Let’s introduce a very simple situation. Let us assume that behind the equation:

T(t) = Known Science + Unknown
T (t) = K(t) + ?(t)

We work out the science and determine:

K(t) = f(t) = 0.16 × t

But when we measure the system we find:

T(t) = 0.48 × t

So both the knowns and the unknowns are increasing with t (in the measured period) in this manner:

T(t) = (0.16 × t) + (0.32 × t)

Now let’s see where the two approaches take us:

Engineer:	There are things we will not know	T(t) = f(t) + ?(t) ?(t) = (0.32 × t)
Academic:	We must know everything so we must make it fit	T(t) = 4.8 × f(t)

So the engineer/skeptic who is willing to admit how much they do not know says there’s a huge thing they give the name of “natural variation”. But the academic (who has a culture that they must know everything) assumes that everything must be known so on the basis “if it’s not f(t) what is it?” they end up scaling f(t) to fit.

But this will never happen in practice!!

The problem is it can happen and it has happened in the case of the global warming models.
The graph to the left shows how the “unequivocal” science of CO2 warming, the direct and quantifiable effect of CO2 of about 1C warming for a doubling of CO2, in neither shape or form, matches the much larger actual global temperature variation seen.
It is very clear that the “knowns” (i.e. science) are a minuscule part of the big picture and that precisely what I have described has taken place.
But you will not have seen this graph before (except on the SCEF website).
This is where the statement we must “ know what happened / is happening“, leads. It has led academics researching climate up a blind alleyway of ascribing all the change to what little they know.
This is why the engineering philosophy of “there will always be things we do not know – we just have to live with them” is so effective. The two approaches lead to two fundamentally different predictions:

	Belief	Prediction
Academic:	All the rise from 1970-2000 “must be explained” so due to CO2.	The rise was due to CO2 so temperatures will keep rising {and worse} up to 6C/century
Engineer:	We know CO2 is a greenhouse gas but there is also much we do not know that we call “natural variation”	The best estimate based on the science, is 1C/century. But Natural variation could make this smaller or larger

And the what is the result when we get a pause?

	Assertion	Response to the pause
Academic:	The science is “unequivocal”	The pause does not exist
Engineer:	Modest warming is easily swamped by natural variation.	As expected, CO2 warming has been swamped by natural variation and as expected we have a pause.