Yes, Katharina, analogies and metaphors are (mental) maps.
Indeed, a geographic map is a metaphor of the underlying physical reality. I like your metaphor for metaphors, and I’ll use it in my reply.
But first a clarifying point. I never made a principle point against analogies and metaphors. I’m always against principles – they are not “truths” but metaphors of reality and mental constructs.
The first paragraph of my blog 137 is effusive of compliments for analogies – as long they closely track material reality, i.e. have many “points of congruence”. Where I raise warning flags is where the maps track mental maps rather than “geography”. If the mental map is flawed, the metaphor will be as well. Metaphors are bundle packets of “pre-compressed knowledge” from a different context: it’s an “all or nothing” proposition. By smuggling in spurious elements metaphors can seriously mislead by giving us the cuddly feel that we know and understand. A distortion of a distortion seldom yields undistorted understanding.
Ideally, mapping should allow for “one on one” congruence. A point on a map univocally identifies a point in the landscape. By optimizing this process over time we get to a situation where there is ONE map for ONE landscape. If we have MANY maps for one landscape, or if ONE map describes MANY landscapes, we know something is wrong. Expect frantic frowning warnings from me.
A map, in any case, is never a one on one match of reality. We need to compress the information. So we leave 99% of the reality out of the map, and only include say, physical features like heights and waters. It is a convention. As long as we know – it’s fine, provided we know and remember. Just let’s not use a surface map to identify oil deposits or wind directions.
Maps not only convey facts on the ground, inevitably they also include mental constructs we have about the underlying reality. The “myth of continents” is a good example. In everyday parlance Europe is a continent – geographically, this is nonsense, it is a peninsula. Politically, the “Orient” began in Morocco for a while – how odd for a country to the west of Europe. The standard Mercator projection map overstates the size of certain geographical areas. Mental constructs that overlay maps make them inaccurate, nay misleading. These hidden deletions and interpretations are seldom innocent. At best they are part of “silent knowledge” we share. More critically, maps are ideological statements, which organize reality in accordance to our political intents.
Maps finally depend on their use. The metaphor of the atom as astronomical system is useful to teach chemistry and is still valid in this context. If one wants to understand the fundamental forces of the universe, however, the canonical metaphor is that of ice skaters exchanging particles like the Higgs Boson on the go. It shows how a quantum of force moves back and forth.
I’ve given up on “the truth” long ago. The only thing that counts is usefulness of our ignorance in the immediate context – consequentialism is the philosophical term. All maps are ideological statements – this is inevitable, given the structure of brains. The question is, how much ideology do we tolerate? This requires that we turn a metaphor inside out and ask ourselves: what does it say, and what does it hide? Worse, what does it suggest silently? And what are the likely (and unexpected) consequences?
Let’s apply now what I’ve said to your “house on fire” metaphor. The first thing one notes is that it is a generic description. It is but one of numerous metaphors one could use to convey the essentially same message – urgency (no content is involved). It could be avalanche, nuclear explosion, whatever else comes handy. In mathematics (an analogy) we would speak of over-determination: too many variables chase too few equations – the result is vagueness. One could use other metaphors equally well, or use the same metaphor in other situations – for instance in the current Middle East context. Or to describe Iran’s alleged quest for nuclear capability. The “house on fire” generic suitability betrays its high rhetorical content.
Analyzing its fit to climate change more closely, one can observe that the metaphor (a) conveys urgency which may be exaggerated; (b) suggests knowledge of the evolving event we do not have, (c) intimates solutions which, on closer inspection, may prove to be less than appropriate; (d) evokes the precautionary principle that we should act, rather than think, or adapt hoping for the best. It has “true and present danger” written all over. The (un)intended consequence is rush “to do something, anything”. We did. The policy choice turned out to be another analogy – scaling up from the Montreal Protocol. It was the frog trying to be a bull (this analogy is apt; it is cautionary, rather than persuasive). The outcome – foreseeable – was stagnation.
Robert Frost said: “poetry is what gets lost in translation”. Analogies are translations. Analogies compress the infinite reality into information packages we can use to survive for a while (not to master reality, that’s hubris). Most of the information is lost, and extraneous gets smuggled in – as when you have a biased translator, or when the word’s aura differs from the original.
Our brain has at least two “layers” – the emotional and the rational. Emotions are fast; rationality is slow and entails hard thinking. Rhetoric appeals to emotions and tendentially shortcuts rationality. Depending on the rhetorical content an analogy can either be a hindrance or an adjuvant.
I’ll make a final analogy with catalysts. “Catalysis is the change in rate of a chemical reaction due to the participation of a substance called a catalyst. Unlike other reagents that participate in the chemical reaction, a catalyst is not consumed by the reaction itself.” If an analogy accelerates the thinking process – it is acceptable. If it distorts it – it is unacceptable. We need to verify – and not just rely on our cuddly feeling.
 The metaphor here is: “if it walks like a duck and quacks like a duck… it is a duck” – which is an unwarranted inference. A single point of incongruence may destroy the analogy, so it behoves us to verify that we don’t think like a duck.
 Martin W. LEWIS – Karen E. WIGEN (1997): The myth of continents. A critique of metageography. California University Press, Berkeley.
 The Mercator projection portrays Greenland as larger than Australia; in actuality, Australia is more than three and a half times larger than Greenland. http://en.wikipedia.org/wiki/Mercator_projection