Beware of phantom metaphors

Of undead trees, silent genes and chaperone proteins

Keith S. Taber

These zombie metaphors become (like a neutron star) 'undead' as they pass from the expert's text to the novice's mind. They are phantom metaphors in the sense that they will manifest as 'living' metaphors to the uninitiated even though the expert user knows they have been put to death.

…the novice or non-specialist has no way of knowing what is the refined meaning and what is just semantic residue.

I have become a little obsessed with the figurative language used to explain science. Science often involves quite abstract ideas, which – by definition, being abstract – do not directly relate to familiar concrete objects and experiences. Learning theory suggests that to make good sense of new information, we need to relate it existing mental resources (existing knowledge and understanding; familiar experiences or images, and so forth).

This implies a paradox (indeed this is related to a traditional puzzle known as 'the learning paradox'):

we can only make sense of things we can relate to in terms of past experience
the science curriculum sets out a large number of abstract ideas that do not directly relate to the everyday experience of most people

We are all familiar with green plants, and may know from practical experience that they need light and moisture, but that direct everyday, phenomenal, experience is some way from the abstract concept of photosynthesis. This point could be repeated regarding any number of other ideas met in science courses: magnetic hysteresis, p-orbitals, electron spin, genomes, metabolism, uniform electrical fields, electronegativity…

Now, perhaps any science teachers or scientists who read that passage may feel I am exaggerating – they can no doubt readily bring to mind images representing hysteresis and fields and genomes, and equations for photosynthesis with chemical formulae, and the values electron spin can take (±¹/₂, obviously). These things will be familiar and can be readily represented in 'working memory' (where we undertake deliberate thinking), so to be applied or mentipulated in various ways. But that is a result of the familiarity of expertise built up over a good deal of time. Sure, I can bring to mind a representation of a double bond or a methane molecule or the earth's magnetic field as easily as I can bring to mind an image of a table or a bus or a blackbird. This is useful for a science educator, but is also a potential barrier to putting oneself in the place of a novice learner.

A key is that "the science curriculum sets out a large number of abstract ideas that do not directly relate to the everyday experience of most people". And teachers, and other science communicators (such as journalists and science writers) can address this in two ways.

The best response, when possible, is to provide experiences (through demonstrations and practical activities) that motivate the concepts to be learnt. By motivate, I mean that this experience provides a recognised need for the explanations (as well as associated technical terminology) to make sense of the experiences. Practical work in science classes can be used in various ways, and rather than teach students about some theory, and then demonstrate it, it may be possible to offer experiences which raise questions and wonderment that will give the explanations 'epistemic relevance' (Taber, 2015). The learner will not just be learning about concept X because it is in a syllabus, but because they want to know why Y happened. Now that may seem idealistic – but most children start curious (perhaps before the routine nature of formal education somewhat dulls this) and it is something to aim for.

But of course some things are too slow, too fast, too big, or too small (or too dangerous or too expensive) to bring into the classroom. One cannot* teach the big bang by giving learners a direct experience which will lead to them asking questions that can be satisfactory answered by introducing the canonical scientific account. (* Perhaps I am wrong – if so, I would like to see ther lesson plan.)

Tools for making the unfamiliar familiar

So, the other approach to 'making the unfamiliar familiar' needs to be indirect, perhaps with videos and simulations and models which represent the inaccessible experiences – supported by (and where those tools are not available, through) a narrative where the teacher talks new entities into existence in a learner's 'mind (Lemke, 1990).

Important tools here are analogies where the learner is told that the unknown 'X' is in some ways a bit like the very familiar 'A'. There are a great many examples of analogies used in explaining science. Here are just a few:

(Many more examples of analogies can be hound here)

Now analogies (like models more generally) are never perfect. X is like A in some ways, but in other ways X is not at all like A. (Otherwise, an X would be an A, and so no more familiar than what is being introduced.) This imperfect mapping does not matter because the use of analogy is not just (i) saying 'X' is in some ways a bit like 'A', as having established that anchor in the learner's prior experience, the teacher develops the comparison by exploring with the learners (ii) the ways in which the two things are alike and (iii) the ways they are not alike, and so starts to build up the learner's familiarity with the nature and properties of X.

"…for effective use of teaching analogies:

carefully analyse the analogy in advance and be confident that the analogy, and, in particular, the features of the positive analogy that are useful for teaching, are indeed already familiar to learners in the class;

be explicit about the use of the analogy as a tool, a kind of model or device for generating conjectures to think about;

be very explicit about the structural features being mapped across, so it is very clear which features of the analogue are being drawn upon to introduce the target knowledge…

explore aspects of the negative analogy that could mislead learners (perhaps invite learners to consider other features of the analogue and suggest aspects that may or may not transfer);

consider the analogy as part of a scaffolding strategy – an interim support to be withdrawn as soon as it is no longer needed as learners are comfortable with the target concept."

(Taber, 2024)

A weaker technique than analogy is simile: simply pointing out that X is like A. This is clearly not going to do the work of an analogy, as when introducing a whole new theoretical concept, but has a role 'in passing' when pointing out some single feature or function.

Simile is widely used in communicating science. There are descriptive similes that tell us that something unfamiliar physically resembles something familiar ('lacework', 'bristle-like', 'like a boat') : this technique was widely used by naturalists in describing things they observed, such as novel species, and was especially valuable before the invention of photography. Contemporary science communicators also commonly make use of this technique with more abstract comparisons to functions and properties rather than just appearance:

(Many more examples can be found here.)

Metaphorical mystery

Now metaphor is like simile, except that the comparison is implicit. That is, consider the difference between saying:

a mitochondrion is like the engine room of a cell; and
a mitochondrion is the engine room of a cell;

As in the simile, the user does not go on to explain how the mitochondria may be understood in this way (which would constitute an analogy) and so the audience is required to do some work (so similes should only be used in teaching when the teacher is confident meanings are obvious to the learners). But with the metaphor the audience has to first even recognise there is a comparison being made, as this is not explicit. After all,the following two propositions have parallel structures:

'a mitochondrion is the engine room of a cell'
'a headteacher is the professional leader of a school staff'

In one case identity is intended (a headteacher IS the professional leader of a school staff), but in the other case there is only a figurative identity: a mitochondrion is not an engine room (even if that could be the basis of an analogy that could be productively explored). So, I advise teachers to avoid metaphor in their explanations, and to always make it clear they are using a comparison. It may seem obvious that a tiny organelle is not (and cannot be) the same thing as the engine room in a ship; but why add to the learner's task in making sense of teaching by adding the need for an extra stage of interpretation that could be avoided?

Manifold metaphors

That said, metaphors are very common in science communication. Here are just a few examples of many I have collected.

(Many more examples can be found here.)

Perhaps we should not be surprised at metaphors being so ubiquitous because metaphor is a core feature of language. They are so commonplace that we do not always consciously notice them, but can often simply read or listen straight past them. Even if we notice there is a metaphor in a text, where it is successful we immediately grasp the meaning and so it aids understanding rather than confounding it. I am hoping that my use of the metaphor 'anchor', above, worked that way. You may have spotted it was a metaphor – but I hope you did not have to stop reading and puzzle out what I meant by it in that context.

An anchor (Image by Tanya from Pixabay) but what has this got to do with meaningful learning?

In particular, language often develops by metaphor. So a term that is used initially as a metaphor, sometimes get taken up and repeated to such an extent that some decades later it is treated as a conventional meaning for a term and no longer considered a metaphor. Thus the language grows. So 'charge', in 'electrical charge', was initially a metaphor, an attempt to describe something new in terms of something already familiar (the charge that needed to be placed in a firearm ready for the next shot) but is not considered so now. Sometimes the 'new' meaning comes to exist alongside the original as a kind of homonym (as separate meanings – as with the word 'bank' when referring to a river bank and a financial institution), and sometimes the original meaning falls out of use (as few people use firearms today, and even fewer charge them with shot and gun powder before use).

So, terms that are at one time metaphorical can become 'literal' over time, and these are sometimes called dead metaphors. They are also known as historical or frozen metaphors. The latter term appeals (although it is a metaphor, of course! – words do not actually freeze) because it suggests a change of state that may take some time. That is, there are active metaphors, and frozen metaphors, and then some 'freezing metaphors' that are beginning to be widely understood directly without being understood as figurative, but where this transformation is not yet complete.

I am sure there are plenty of terms that are in common use in the language where, if people were asked, some, but not all, would recognise them as metaphorical (dying metaphors? freezing metaphors?) – and where perhaps decade-on-decade repeat surveys would show some of these had died/frozen, while new metaphors were appearing, becoming widely used, and slowly starting to solidify.

At the risk of pushing an analogy too far, we might note that the state of a sample of a substance depends on the conditions (there are no ice sheets over the Caribbean islands), so if we extend this freezing metaphor, might we find metaphors that have frozen in some environments but are still fluid in other conditions?

Zombie metaphors?

Actually, I think this is likely very common in technical fields like the sciences. I have written here about some of the language used by astronomers when discussing the births, life-cycles and deaths of stars.

The passing of stars (Birth, death, and afterlife in the universe)

The complicated social lives of stars – Stealing, escaping, and blowing-off in space

Clearly these terms were introduced metaphorically. But now they are treated as if technical terms – so, now, stars really do get born, and really do die because these terms now refer to what actually happens to stars, rather than just to processes that had some similarity to what happened to stars.

I think this is potentially problematic from an educational perspective, as the novice who reads a popular astronomy book or listens to a podcast or hears a news report where stars are said to be born, live out their long lives, and die, is unfamiliar with the astronomical processes labelled in this way, and can only understand these terms metaphorically by reference to how familiar living [sic, non-figuratively living] things are born, live, and die. A pet dog that dies is no longer around, but a large star that 'dies' in a supernova explosion may then live on as a neutron star – a bit like some phoenix that rises from the funeral ashes to be reborn.

This is a mosaic image, one of the largest ever taken by NASA's Hubble Space Telescope, of the Crab Nebula, a six-light-year-wide expanding remnant of a star's supernova explosion.

Reincarnation? The Crab Nebula as seen by the Hubble Space Telescope (HST). The Nebula is a Supernova Nebula — One formed from a supernova which left a millisecond pulsar at its center. So was the explosion the death of as star – or was it just a transition to a new phase of the star's life cycle?

(Source, Wikimedia commons; Original source Hubble images due to NASA, STSci, ESA.)

I am not suggesting that people will be generally confused about heavenly bodies being actually alive (even if for many centuries they were widely assumed to be so – many people seem to have thought stars and planets are living beings like humans), but because – for the experts 'born', 'live', 'die' are no longer metaphors – they may be are used without awareness of how a novice may struggle to fully appreciate their 'technical' implications.

So, in a sense, these metaphors become 'undead' (like the neutron star?) as they pass from the expert's text to the novice's mind. They are phantom metaphors in the sense that they will manifest as 'living' metaphors to the uninitiated even though the expert user knows they have (through habitual use) been put to death.

Not just out of this world…

I suspect that there are zombie metaphors in use not just in astronomy, but in many technical fields. This means that any of us who are reading 'out of specialism' are likely to mistake phantoms for live metaphors even when an author or speaker is using a term non-figuratively with a meaning that has long ago solidified in that specific discourse environment.

When a pure substance freezes it may exclude impurities. So, for example, a sample of sea water will start to freeze, and the ice forming will exclude the salts dissolved in the water (so the salt concentration in the remaining solution increases). When a metaphor freezes to become a technical term it retains the aspect of the comparison that were originally intended figuratively, but not other features that are not relevant – they get 'frozen out' so to speak. The expert has in mind the 'purified' meaning, and does not bring unintended associations to mind. But the non-specialist has no way of knowing what is the refined meaning and what is just semantic residue.

Figuring out erythrocytes…

Consider, for example, a textbook chapter entitled "Anemias, Red Cells, and the Essential Elements of Red Cell Homeostasis" (Benz, 2018). This chapter uses a range of figures of speech to help communicate technical ideas. Some of these can be glossed:

erythrocyte membrane cytoskeleton protein is like a spring (simile)
proteins act like molecular swivels or molecular hinges (similes)
red blood cells have limited arsenals (metaphor)
surface abnormalities on red blood cells are polished away (introduced as a simile, then used as metaphor)

There are also a couple of places where phrasing might be seen to move beyond simple metaphor to anthropomorphism: that is, writing that seems to imply non-sentient entities have preferences and desires or act after conscious deliberation:

The chapter also refers to the proteins known as Ankyrin. This is a technical term of course. A review article relates that

"Ankyrin is a binding protein linking structural proteins of the cytoplasm to spectrin, a protein present in the membrane cytoskeleton in human erythrocytes that functions as an anchoring system to provide resistance to shear stress."

Caputi & Navarra, 2020

Indeed, ankyrin gets it's name from the Greek word for anchor. So ankyrin is not a metaphor, but derives its name metaphorically in relation to its perceived function.

part of the structure of ankyrin proteins

Ribbon diagram of a fragment of the membrane-binding domain of human erythrocytic ankyrin (left-hand image, from Wikipedia commons), member of a class of proteins named after an anchor (right-hand image).

But I also noticed a number of other terms which manifested as metaphors, but which I do not think would be considered metaphors by specialists. In the field, they would be dead metaphors, but to a novice they might appear as phantoms, assumed to be meant metaphorically:

chaperone proteins: assembly of hemoglobin from newly synthesized globin chains requires the presence of chaperone proteins
cross-talk: there is little or no direct 'cross talk' between the α-like and β-like gene clusters
don't eat me signals: membrane protein CD47 is known to be a 'don't eat me' signal
membrane leaflets: phosphatidyl serine is largely confined to the inner leaflet of the lipid bilayer
silent genes: globin genes are completely silent in all other tissues
vascular trees; red cells must be able to resist adherence to the walls of the vascular tree

These can seem to be figures of speech, with the fluid quality of offering the reader the creative act of deciding which properties to transfer across from the metaphor/simile: but actually are all widely used terms in the field, and so actually have definite 'frozen' meanings. A vascular tree has branches (and twigs) but no leaves or fruits.

Perhaps there is not too much potential here to confuse readers (especially given the intended readership for this particular text would be professional / graduate), but it does reinforce the idea that communicating science is a challenge when not only, as is often noted, so much of the language of science texts is technical; but a lot of technical terms are dead metaphors: with frozen meanings that have the potential to melt back to life, and invite more fluid interpretations from learners.

Work cited:

Benz, Edward J. (2018) Anemias, red cells, and the essential elements of red cell homeostasis, in Edward J. Benz, Nancy Berliner, & Fred J. Schiffman, Anemia. Pathophysiology, Diagnosis, and Management, Cambridge University Press, 1-13.
Caputi, Achille Patrizio & Navarra, Pierluigi (2020) Beyond antibodies: ankyrins and DARPins. From basic research to drug approval. Current Opinion in Pharmacology, 51, April 2020, pp.93-101.
Lemke, Jay L. (1990) Talking Science: Language, Learning, and Values, Bloomsbury Academic.
Taber, K. S. (2015) Epistemic relevance and learning chemistry in an academic context. In I. Eilks & A. Hofstein (Eds.), Relevant Chemistry Education: From Theory to Practice (pp. 79-100). Sense Publishers. [Download chapter]
Taber, Keith S. (2024) Chemical pedagogy. Instructional approaches and teaching techniques in chemistry. Royal Society of Chemistry. [Download Chapter 1]

Share on Facebook

Author: Keith

Former school and college science teacher, teacher educator, research supervisor, and research methods lecturer. Emeritus Professor of Science Education at the University of Cambridge. View all posts by Keith