Are metaphors a good thing?

Are metaphors a good thing? Some see metaphors as suspect.

Pro

 * Without metaphors, poetry would be greatly impoverished.
 * This is consistent with metaphors being bad in some way, for some purposes other than poetry.
 * The question is not whether metaphors are bad for something but rather whether they are bad. If they are bad overall, one would perhaps like to ban them or consider their use indecent. And if so, poetry would be banned or be considered indecent.
 * Metaphors help concise naming.
 * Concise naming is not the ultimate or absolute value. One should perhaps aim at less concise yet clearer and less ambiguous naming.
 * If one seriously tries to do all naming without metaphor, one will realize how cumbersome it gets. The supporters of metaphor-free naming should produce a dictionary of terminology that does without metaphor, to show us the feasibility. That has not been done so far.
 * Metaphors cannot be done without: the language is a store of dead metaphors, of things that once were live metaphors before they became part of the standard meaning of a word.
 * We can at least eliminate as much metaphor as we can. If metaphors are bad but necessary things, we may reduce them to the very minimum. Thus, the above is inconclusive.
 * To expand on the above objection, the term "dead metaphor" itself is a poor metaphor. Dead metaphors are much more of fossilized metaphors or ossified ones. Since, a dead thing often decays and disappears, whereas dead metaphor solidifies and becomes like a standard tool use in large frequencies without being worn down any further. This shows how suspicion of metaphor is a good thing.
 * That supports the notion that there are better or worse metaphors, not that metaphors per se are a bad thing. Indeed, the words "fossilized", "ossified" and "solidified" point to interesting comparisons and analogies.
 * Dead metaphor is not a species of metaphor; it is metaphorical use of the word "dead" to act as alienans, e.g. like "false director" is not a director. By becoming "dead" or fossilized, metaphor ceases to be one, like a mollusc ceases to be mollusc by fossilization. Thus, the main objection against metaphor is against live metaphors. Furthermore, even dead metaphors can be eliminated in part, serving the goal of getting rid of metaphors to a large extent, even if not completely.
 * Metaphors are a concise tool of cognition and hypothesis discovery: by asserting A is B when it is in fact not so, we are asking the other party to ask: what properties of B apply to A and what properties of B do not? And by allowing the metaphorical process, we may ask: what is the most concise metaphor for A? Thus, should something be described as gold or rather as a diamond? Both are high in price, but they differ in richness of shape. If something is high in price or value and rich in shape or form, it is better described metaphorically as diamond rather than gold. Thus, the exploration of metaphors helps exploration of the network of notions and properties, and helps guide thought.
 * But for that one does not need metaphor: simile or other literal speech is sufficient. Thus, one may say "A is like B", "A is like B in some regards" or "Some properties of B apply to A".
 * Fair enough. However, the above phrasing is cumbersome.
 * A bit cumbersome but perfectly workable for analytical purposes, robbing the sophists and demagogues of their powerful tool of deception.
 * The insistence on the explicit flagging using "like" or "as" as in a simile is just pedantry, excessive literalism. An intelligent reader or debater can understand the "like" or "as" as implied. The difference between a metaphor and a simile is merely formal, in the flagging.
 * A problem with the above is that it creates a barrier for refutation and criticism of statements and arguments, convenient for deceptive arguers. When presented with a false claim "A is B", the refuter has to refute not only "A is B" but also "A is like B" and perhaps also "A is a little bit like B, being like B in at least one salient regard".
 * A problem with the above is that it allows deceptive arguments depending on switching from a metaphorical reading to a non-metaphorical one. Thus, if the arguer says that "A is war, so we have to respond with war", it sounds at least nominally plausible on the surface. Properly flagged, the statement becomes "A is like war, so we have to respond with war", which is much less forceful on the surface. One of the purposes of articulation of arguments and careful choice of words is to bring things to the surface so that the logical engines of the mind have a much easier work to do. In the analyzed utterance, an honest arguer does not need to omit "like", a single word of a single syllable.
 * The use of "on the surface" is a metaphor.
 * It seems rather to be a dead metaphor, one that has become conventionalized. If it is not, it can be improved by the explicit flagging using "as if", thus, "as if on the surface".
 * For something to be a dead or fossilized metaphor, it first needed to be a live metaphor. If we ban all metaphor, we will have no metaphor and will lose all the benefits.
 * Fair enough. A carefully worded motion would perhaps not ask whether metaphors are "good", but rather whether they are dangerous for being an easy tool of deception, whether a requirement to largely avoid metaphors in arguments is good and practicable, and the like. A complete ban on metaphors is unworkable, but a healthy suspicion of metaphors in arguments and persuasive speech sounds good. When someone uses a metaphor, they need to count on being called out on it, and asked to reformulate the metaphor in more literal language.
 * Metaphors serve for initial expression of things that are hard to express otherwise. Thus, one can speak of throwing sand to someone's eyes to indicate that a certain kind of argument or wording was used to make thinking harder. The metaphor points to something more specific than the generic phrase "make thinking harder".

Con

 * Metaphors are a powerful tool of rhetorical deception, of creation of superficially appealing yet incorrect arguments.
 * If true, that makes them bad for some purposes, but not bad overall.
 * True. However, the discovery of honest and correct argument and prevention of bad argument is of utmost importance. The use of metaphor causes a great harm to the argumentation enterprise. It is at least plausible or worth considering that metaphors do more harm than good overall.
 * "Enterprise"?
 * Metaphor is the mother of all evil, if you will. There you get me: a metaphor and a hyperbole as well.
 * Metaphors serve not only to deceive others but to deceive ourselves in our own reasoning. As most non-literal speech, they raise a bar against criticism and refutation. They make it harder to see what exactly is being claimed. When A is metaphorically claimed to be B, it is not clear which properties of B are claimed to apply to A, and there are usually some properties of B that do apply, which makes metaphors hard to refute. That makes metaphors like sand thrown at the bull's eye (which is a simile, not a metaphor).
 * The criticism applies to simile and comparison as well: when A is claimed to be like B, it is not clear in which regards it is like, and it is usually like in some regards, making it hard to refute. Thus, the criticism is not specific to metaphor.
 * That only means that similes and comparisons are also problematic, not that metaphors are unproblematic.
 * That opens a separate discussion about whether similes, comparisons and analogies are good. It is far from obvious that all such things are bad because they lead to self-deception. For instance, the hydraulic analogy was used during early attempts to understand electricity. An analogy was posited between a macroscopic object such as a billiard ball and an atom in physics. Doing away with all analogies does not seem practicable or wise.