What an Analogy Copies Is Not Appearance but Relation, and a Residue Always Remains

Inside a reader who has just been given an analogy, what happens is not a match of appearances but an application of relations.

The reader calls up the memory they hold of the material offered as the analogy. Inside that memory sit the relations belonging to the material. Coming in, accumulating, leaking. Sowing, waiting, sprouting. Applying those relations to the content being explained — that is the operation running on the reader’s side.

Analogy usually gets explained as replacement. Replace difficult content with something familiar. Bring it close to what the other party knows. The word replacement carries a premise: that the content to be conveyed exists first, and is being dressed in a different appearance. The inside stays the same and only the outside changes.

No change of clothes is taking place. What the reader applies is relation rather than appearance, so what settles whether an analogy is good is not familiarity. It is whether the relations the reader holds and the relations you wish to convey agree. Familiar but relationally different, and the reader will understand accurately, and wrongly.

Three things can be handled in what follows. The first is what the fact that relations rather than attributes get copied changes about the criterion for choosing an analogy. The second is where the surplus implications carried in by familiar material come from. The third is what happens when you write out, rather than leave unsaid, the part that no analogy manages to copy.

The things no answer arrives for go here first as well. How many analogies a single piece may contain, how many lines they may occupy — no figure gets written for these. They vary with the number of steps in the structure being handled and with the quantity of premises the reader already holds.

Where the damage actually shows up belongs here too. A divergence of relation does not appear in the form of something failing to get through. The reader says they understood, and they did understand. The state in which what was understood differs from your intent is harder to detect than the state in which nothing arrived. Failure to get through brings questions; something else getting through brings none.

And the more an analogy diverges, the longer it survives. The response being good, the writer starts using that analogy repeatedly. The more it gets used, the more it becomes the skeleton of your own explanation, and a skeleton drops out of what is under examination.

To state it plainly first: this is not an argument for giving up analogy. As a means of letting a reader hold an image, it has almost no substitute. What gets handled here is what remains after the image has formed.

What Are the People Good at Analogy Said to Be Choosing

The advice offered on choosing an analogy gathers into five items.

First, being simple. If the analogy itself requires a long preamble, or its content is convoluted, you end up explaining the explanation. An analogy harder than the thing it serves does not fulfil its role.

Second, taking the listener’s knowledge into account. Does the other party know that material. Does it get through without prior knowledge. Skip this check and the discussion proceeds with the analogy failing to land.

Third, choosing expressions that reach everyone. Material confined to a particular hobby or industry works powerfully on those who know it and conveys nothing to those who do not.

Fourth, using familiar material. Replace with territories most people have experienced — cooking, vehicles, houses, school — and the other party can draw it towards themselves and think.

Fifth, actually trying it. Build the habit, when something feels poorly conveyed, of considering whether an analogy could simplify it. Say it aloud even if the first attempts fail.

All five are on target. A complicated analogy is useless, and material the other party does not know fails to land. That familiar territories offer more memories to call up is a fact as well. An unfamiliar term replaced by a familiar scene makes reading on easier by that fact alone. The effect of building an entrance is certain.

Underneath the five, one shared way of seeing has been laid down. Analogy is organised as an operation for lowering difficulty. Bring the hard down to the easy. A one-directional conversion, from above to below.

Treat it purely as a matter of difficulty and the discussion proceeds on the premise that being made easy is sufficient for conveying correctly. The premise is that the content being conveyed does not change.

Sometimes the content has changed by exactly the amount it was made easy. Where the relations held by the substituted material differ from the relations of the original content, the reader receives something easy, and different. The result something else got conveyed, as opposed to nothing got conveyed, cannot be detected by any measure of ease.

The same way of seeing reaches how skill gets measured. Whether an analogy is any good ends up measured by how fast the listener responds. A fast response means the analogy struck the reader’s memory hard, and whether the memory it struck held the right relations does not appear in the speed.

As a result, the direction of improvement turns towards stock of material. Increase the range of usable material, hold a reserve of familiar scenes. More stock raises the odds of striking something, and what travels along with the strike has nothing to do with the size of the stock.

Easy to Understand and Structurally the Same Are Different Things

Two criteria have been mixed together. Ease of understanding is a matter of the reader’s load; correctness is a matter of relations agreeing.

To explain a state of flat sales, somebody says it is like pouring water into a bucket with a hole in it. It is an easy analogy. Bucket, hole, and water are known to everyone.

The relation this analogy carries is that there is a quantity coming in and a quantity being lost, and nothing accumulates unless the quantity lost is reduced. If the content to be explained is departures of existing customers exceed acquisition of new ones, the relations agree.

Consider the case where they do not. Suppose the content to be explained is what you are pouring into is not your bucket but somebody else’s container. The bucket analogy now turns the reader towards the hole. The reader starts considering how to reduce the departure rate. The question of whose container it is does not exist inside the analogy, so it never arises.

Lean towards just use a different analogy in that case and the adjudication itself drops out. The problem is that if only is it easy to understand gets examined at the moment of choosing, this disagreement never gets adjudicated. A bucket with a hole is easy in both cases. The criterion of ease cannot separate the two.

The degree of agreement worth pursuing has a ceiling. Demand too much agreement and analogy becomes unusable. What to require is not complete agreement but that the relation you wish to convey agrees, and that the writer knows which parts do not. The latter is what will be called grasping the residue.

The easier the analogy, the less correction works. An image stands firmly inside the reader, so adding not quite, rather afterwards leaves the first image in place. A weaker analogy can be corrected later.

That upper bound carries a condition on the other side too. The writer knowing which parts fail to agree means nothing if it never reaches the reader. A residue that is merely known is reassurance for the writer and nothing else. The procedure for putting it into a form that can be handed over is to write the residue out first.

And this disagreement does not appear in the response either. The reader reports understanding, and the report is accurate. Failure to get through brings questions. When no question arrives and behaviour diverges from your intent, that is the moment to suspect the disagreement. A scarcity of questions is not an indicator that something got through.

What Is Copied Is Not Appearance but Relation

Why do analogies whose appearance resembles the subject sometimes fail. Because what gets copied is not appearance.

Gentner, in 1983, formalised the position that what gets mapped in metaphor and analogy is not the attributes of the objects but the relations between them (Gentner, 1983, Cognitive Science, 7(2), 155–170). The correspondence between the atom and the solar system aids understanding not because spheres resemble one another but because a centre, and things orbiting it, is a relation the two share.

This paper presents a theoretical framework accounting for the processing of analogy; it is not a report measuring the effect of analogies in practical settings. What can be taken from it is the direction in which the criterion for choosing moves from attribute to relation.

Once the direction moves, the work changes. Instead of examining whether things resemble one another, you examine whether the sequence of relations is the same.

Back at the bucket, a bucket and a business look nothing alike. They connect anyway because the sequence enters, accumulates, leaks is shared. Conversely, an analogy taken from a territory close in appearance fails when the relations differ.

The practical procedure is simple. Write the content you wish to convey out in the form of relations first. A settles B, B is invisible, and what is visible is C, in that manner. Then look for scenes holding the same sequence of relations.

The sense that a good analogy simply comes to mind is real enough. Sometimes it does come. Whether the one that came is correct, however, cannot be adjudicated without following this procedure. Analogies that come readily are often analogies already in circulation, and being in circulation means having been built for different content.

The degree of agreement has a range too. There are settings where two out of three elements corresponding is enough, and settings where one element out of place breeds misunderstanding. The adjudication is made on whether the one element out of place changes the reader’s behaviour.

This formalisation carries a further implication. Relations being what gets copied means material from a distant field is sometimes the more effective choice. Where the fields are close, the reader begins comparing attributes rather than relations.

The atom and the solar system work as a strong example partly because not one of their attributes resembles the other’s. Nothing resembling anything, the only thing shared is relation. The reader ends up extracting relation alone.

Attempting to write the content out in the form of relations sometimes fails. This is not an inability to find an analogy but an indication that the content you wish to convey has not itself been organised as relations. Here, the work of looking for an analogy cannot proceed.

Familiar Material Sometimes Manufactures Misunderstanding Instead

Familiarity has an advantage and a disadvantage of the same size.

The advantage is that there are more memories to call up. Cooking and moving house are things most people have genuinely experienced, so an image stands up readily.

The disadvantage is that too many memories get called up. The writer wanted one relation, and the reader’s memory contains the whole of their experience with that material. Parts the writer never intended come up along with it.

Suppose the launch of a business gets likened to sowing a seed and raising it. Suppose the relation you wanted to carry was nothing more than there is a lag.

The gardening memory the reader calls up contains other elements. That what is sown will eventually sprout. That care makes things grow. That there is a season for it. These travel along too.

As a result, the implication keep at it and it will eventually bear fruit gets conveyed although the writer never said it. The information that some things never bear fruit ranks low inside a gardening memory.

A doubt arises as to whether readers take things in at that granularity. They do not. What gets taken in is the overall atmosphere, and the implications ride on the atmosphere. What arrives explicitly gets checked; what arrives as atmosphere does not. That is why the latter is what survives.

There is no need to avoid familiar material. With no memories to call up, no image stands in the first place. What requires handling is the surplus implication familiar material brings in.

The more familiar the material, the more unconsciously the writer selected it. It gets used with the reason for its selection unexplained, which is why what travelled along with it goes unnoticed as well.

A second problem sits alongside it. The extent of familiarity gets settled by the writer’s own life. Things familiar to you feel familiar to the reader too. This estimate frequently goes wrong.

The adjudication cannot be made on whether you can talk about the material without explanation. You can, because you know it. What can be confirmed is only this: whether the material gets experienced independently of the reader’s attributes. Commuting, paying, and waiting are close to that; child-rearing, driving, and playing an instrument are not.

And getting it wrong produces no symptom. Readers for whom the material did not land leave without asking. The fact of their leaving does not come back to the writer, and not coming back, the material gets used again in the next piece. Material that missed stays in the stock, still missing.

An Analogy Always Leaves a Part That Is Not Copied

An analogy agreeing completely does not exist. If it agreed, it would not be an analogy but the same thing.

Every analogy therefore necessarily leaves a part that does not correspond. The problem is not the remainder itself but the writer not grasping where the remainder sits.

In the gardening example, the part that does not correspond is what is sown basically sprouts. In business, not sprouting is not unusual. Correspondence fails there.

In the bucket example, the part that does not correspond is the bucket belongs to you. Where the publishing happens on a platform, the container is not yours. That is where correspondence fails.

The method for identifying the residue is simple. List four facts that hold of the analogy itself, and confirm one by one whether the same holds of the content you want to explain. What does not hold is the residue.

A worry arises that the labour of doing this exceeds the labour it saves. Writing it out once is enough. If the same analogy gets used repeatedly, the confirmation happens on the first occasion only.

Some residues are harmless. Where the material is one the reader pays no attention to, no damage results. The adjudication is whether the residue changes the reader’s judgement. If it does not, it is a residue that may be left alone.

Write the list from the side of the analogy. Written from the side of the thing being explained, only the items you already believed you had matched come out. The purpose is to catch the elements contained in the analogy that you were not conscious of.

Identifying the residue sometimes makes that analogy unusable — where the central relation itself turns out to be out of place. That is not a loss but a misunderstanding prevented before use.

The writing-out has an upper bound as well. Residues can in principle be listed without end, so a line gets drawn somewhere around four. The basis for the line is one thing: how far the reader attends to that material. In the gardening example there is no need to reach soil acidity.

The claim sometimes changes during the writing-out. With four residues laid side by side, one of them occasionally lands in the position this is in fact what I wanted to say.

That is content getting corrected from the side of the analogy — analogy operating as an instrument of thought rather than an aid to understanding. This operation occurs only where the residue has been written out explicitly, however. Left unwritten, the residue stays outside the writer’s field of view.

Write the Residue First and the Analogy Becomes a Tool

Having grasped the residue, the writer has two options. Say nothing, or write it out.

Writing it out is the stronger move in most cases. State this analogy is the same up to here, and this one point differs, and the reader turns their attention to that one point. A point that has received attention stays in memory.

Return to the bucket. Having used it, you continue: in business, though, that bucket is not necessarily standing on your own land. With that one sentence, the centre of the discussion moves from leakage to ownership. The analogy has operated as a stepping stone for the claim.

Stating the residue is not an admission that the analogy is weak. Showing the limit of the analogy is, directly, showing where the claim sits.

The worry that stating a limit costs persuasive force carries a condition. What loses force is the case where the analogy is being used as grounds. An analogy cannot be grounds. The fact that relations resemble one another is not proof that the content is correct. The role of an analogy is to let the reader hold an image, not to support the claim.

Where it is not being used as grounds, making the limit explicit loses nothing. If something is lost, that means it was being used as grounds.

Nor does every residue need writing. What gets written is only the residue that changes the reader’s judgement. Write all the rest and the notes run longer than the analogy.

Acquire the habit of writing residues and the stage of choosing changes. Knowing you will write the residue later, you start looking for analogies with fewer residues from the outset. That is not something done deliberately; it happens as a result.

The relationship with the reader changes too. Disclosing the limit in advance is a form that hands the reader material for judgement. Handed that material, they can adjudicate for themselves how far the analogy carries into their own situation.

Disclosure has a price. The analogy’s single blow weakens by exactly the amount the limit is written, because the image standing inside the reader becomes an image with a reservation attached. The choice is between handing over a strong image and handing over an image that can be adjudicated. In writing that leaves the judgement with the reader, the latter gets chosen.

Hand it over with the limit concealed and this adjudication cannot be made. The reader carries the analogy off as it stands, uses it in a setting where it does not apply, and arrives at a state where the reason it failed is unavailable. What has been lost here is room for judgement on the reader’s side. An analogy is an instrument for handing over an image, not an instrument for handing down a judgement in the reader’s place.

What Fails to Come to Mind Is Not the Analogy but the Form of the Relation

What is missing in the state no good analogy comes to mind is not stock of material. It is the condition for the search.

The condition is the content you wish to convey, put into the form of a relation. Once the form A settles B has been fixed, scenes with the same form can be searched for.

With the form unfixed, there is nothing to search with, because what to look for a resemblance to has not been determined. Search for an analogy in this state and you end up applying whatever came to mind first from the side of the material.

Applying it can fail. The content gets pulled towards the relations held by the material. The result is that partway through writing, you are saying something other than what you set out to say.

The prescription is to restore the order. Before looking for an analogy, write the content you wish to convey as a relation, in one line. Manage that and the search becomes concrete work. Fail to manage it and the problem is not the analogy, so you return to organising the content.

Some content cannot be put into the form of a relation: where the sensation itself is what you wish to convey, or a particular texture. What gets used there is depiction rather than analogy, and what it requires is not agreement of relations but the reader holding the same kind of memory.

Some cases proceed without a single line. Structure genuinely does come into view while writing. In those cases, however, the analogy gets fitted after the writing is finished.

Once written as a relation, the analogy sometimes turns out to be unnecessary. If one line says it, writing that line gets it through. An analogy is required where the relation spans several steps and cannot be laid out in one line.

The search narrows as well. With the condition fixed, unsuitable material falls away early. The work of weighing one idea at a time becomes the work of looking for something meeting a condition.

And with the condition fixed, the decision to give up when nothing comes becomes available. If a search finds nothing, that relation cannot be carried by analogy. The move to make here is not forcing an analogy on but decomposing the relation and writing it step by step. A relation that can be written decomposed gets through without any analogy.

This decision has an advantage on the cost side too. Time spent searching for an analogy that will not come has no visible upper bound. With the condition fixed, the point at which to break off the search is determined. What remains after breaking off is the other move: writing it step by step.

Overuse Moves the Subject to the Side of the Analogy

Analogy has a problem of quantity. Stack several analogies into one explanation and the reader’s attention moves to the side of the material.

Inside a single explanation, a gardening analogy, a bucket analogy, and a building analogy arrive in sequence. The reader ends up calling up memories of three territories in turn. Calling up carries load, and by the amount of that load, attention to the original content falls.

The three analogies also fail to cohere with one another. Gardening carries a lag, the bucket carries leakage, the building carries sequence. The three never get integrated into one image inside the reader.

What remains, as a result, is the impression that the explanation was thorough, and fragments of content. The impression stays; the relations do not.

Lean towards several aspects require several analogies and a line can still be drawn. Varying the analogy by aspect is a correct decision. The problem is stacking analogies in order to explain one and the same relation. Producing a second because the first did not land is this case.

What is required when something did not land is not a different analogy. It is identifying where the first analogy was out of place and repairing that, or removing the analogy and writing directly. Stack them and there are two places out of place.

Referring back to the same analogy repeatedly within a piece is not stacking. Taking one analogy as an axis and loading several points onto it preserves the relations instead.

There are cases where stacking is legitimate: laying the same relation out for different layers of reader. There, the fact of laying them out gets stated. Write to put it another way and the reader can take up whichever of the two they prefer.

Whether you are stacking is hard for the writer to adjudicate, because the reader’s load does not come back to your side. What can be adjudicated is the form: where two or more materials strike one and the same relation, you are stacking. If there really are two aspects, two relations should have been written out.

An analogy too strong gets remembered in place of the content. When a reader explains the material to somebody later, they sometimes reproduce the analogy alone and cannot reproduce the original claim. Here, what got through was the material rather than the content.

The same thing happens on the writer’s side. Build one strong analogy and subsequent explanation starts getting assembled inside that analogy’s frame. Points outside the frame become awkward to handle — the claim shrinking to fit the analogy, which inverts the order you would want. This inversion cannot be found by reading what got written, because the points you never wrote do not appear on the page.

Reading Where the Analogy Sends the Reader, Before You Publish It

There are three ways to confirm, and each of them finishes before the analogy goes out.

The first confirmation is to write out four things that hold on the side of the analogy and see whether they hold on the side of the thing being explained.

What did not hold is the residue. If it comes to zero, this is not an analogy but the same thing restated. Three or more, and the central relation may not be in agreement either. One or two is the manageable range.

The second confirmation is to see whether the same claim can be written with the analogy removed.

If it can, the analogy is operating as an aid to understanding. If it cannot, the analogy is being used as grounds for the claim. That is a precarious state: the claim collapses the moment the reader declines the analogy.

The third confirmation is to write out the action the reader is likely to take from that analogy.

With the bucket, the reader will try to plug the leak. See whether that matches what you wanted to say. If it does not, the analogy is sending the reader in a different direction. This confirmation looks at the implications the analogy holds rather than at the content the writer intended, which gives it the most detection power of the three.

Ease of understanding is not what any of the three examines. Ease is a condition for an analogy functioning, not evidence that it functioned.

There is a place the three do not reach. None of them can confirm the contents of the memory the reader holds. Experience of the same material differs between people. That part becomes available only from responses after the piece has actually been read.

A figure for how many analogies a piece may hold, or how many lines they may take, could be written only if the number of steps in the relation being handled were fixed. It is not. The number of steps moves with the subject, and it moves with the quantity of premises the reader already holds. Hand over a figure and the allocation gets settled without those conditions being examined.

An analogy is not a change of clothes — that is the one point this began from. What a reader does is apply the relations inside their own memory. If what gets applied is relation, then the criterion for choosing is the sequence of relations rather than familiarity; sequences never agree completely, so a residue remains; and the residue sits in a position that changes the reader’s judgement. The three are not separate matters lined up in order. They are consequences following from the single fact that application is what is taking place.

So there is one procedure to run first on the analogy in front of you. Write out four things that hold on the side of the analogy, and see whether they hold on the side of the thing being explained. The items left over are what your analogy is carrying to the reader in surplus.

How to handle your own experience as material for analogy is treated in how to write primary information, and the method of defining a concept and handing it over in the order in which explanation improves. How a whole piece containing analogies gets assembled is taken up by the principles of article structure. A residue always remaining is not a defect peculiar to analogy. That there is a ceiling on what can be restated at all is taken up in verbalisation as a whole.