A domain with a low ceiling
Keith S. Taber
Organisations use advertising to make claims about their products and services – but how are we meant to understand such claims? (I wonder in particular how people on the autism spectrum are expected to make sense of them.)
Knowledge claims about knowledge claims
As someone who has been professionally concerned with how we understand 'knowledge' (what is it?, what counts as knowledge?) and how we identify it (can it be measured?, can we ever be sure quite what someone else knows?) I've recently been rather irritated by some rather blatantly naive claims made in adverts that I have been repeatedly exposed to when settling down for some relaxing evening viewing. These were not about a product as such, but about the apparently exceptional expertise of the company staff.
Misleading advertising?
I live in a state where there is an advertising standards authority (the ASA) to which consumers can make complaints, and which can take action on misleading advertising.
Here are some selected points from the regulator's website (accessed 21st May, 2022) regarding misleading advertisements:
• The ASA will take into account the impression created by advertisements as well as specific claims. It will rule on the basis of the likely effect on consumers, not the advertiser's intentions.
• Advertisements must not materially mislead or be likely to do so.
• Advertisements must not mislead consumers by omitting material information. They must not mislead by hiding material information or presenting it in an unclear, unintelligible, ambiguous or untimely manner.
• Subjective claims must not mislead the audience; advertisements must not imply that expressions of opinion are objective claims.
• Advertisements must state significant limitations and qualifications.
• Advertisements must not mislead by exaggerating the capability or performance of a product or service.
• Advertisements must not suggest that their claims are universally accepted if a significant division of informed or scientific opinion exists.
https://www.asa.org.uk/type/broadcast/code_section/03.html
Given these points, I feel that some rather specific and unequivocal claims made by an organisation called 'Carpetright' are extremely dubious.
However, I should also note another significant statement which potentially undermines the ASA's ability to apply such principles strictly:
"Obvious exaggerations ('puffery') and claims that the average consumer who sees the advertisement is unlikely to take literally are allowed provided they do not materially mislead."
https://www.asa.org.uk/type/broadcast/code_section/03.html
So, this seems to suggest that claims can be made which are not actually true as long as the 'average consumer' will recognise them as not meant to be trusted and believed. Hm.
At one level, this makes sense. If an advertiser hires an actor to dress up in aluminium foil and announce "I am Thor, God of Thunder, and I will bring my wrath upon you all, except those wearing Taber raincoats who are protected from my powers" then clearly they are not expecting the audience to believe they are seeing a supernatural entity with Godly powers of destruction from which they can be protected by buying the right rainwear. They are expected to interpret this along the lines "our raincoats are pretty effective at keeping you dry when it is raining". As there is no intent to deceive, and as a reasonable person (perhaps that 'average consumer') is unlikely to perceive this as an objective knowledge claim, then this is no more a lie than when an actor plays a part in a fictional play or film.
But I am worried about this 'get-out clause' as it potentially offers a defence for all kinds of claims that cannot be objectively supported.
Who knows the most?
I have been seeing the particularly irritating advertisements regularly for some time now, and had been so grated by the claims in them – in particular by the extremely implausible nature of them when considered collectively – that when I sat down to watch a programme yesterday I actually made a note of the claims in a succession of small sponsor's advertisements positioned at the start and end of each commercial break. (I usually record programmes and fast-forward through such breaks, but these 'sponsoring' adverts are directly segued to the programme content so hard to completely avoid. That said, when you do not use the fast-forward they do offer an obvious indication when to press the mute button as the programme has paused or is just about to recommence.) During the programme the set of claims were cycled through twice (and one of the claims was slightly different in nature to the others and I will leave aside).
The first commercial break was announced with the claim that
"no one knows more about floors than Jeff"
Claim made by 'Careptright'
This is presented as an unqualified and objective statement.
For anyone who has undertaken research to explore the knowledge of learners, this raises a whole array of questions.
What is meant by knowledge in this claim?
Traditionally knowledge is meant to be true justified belief. That is, Jeff has knowledge about floors to the extent that he has beliefs about floors that are are both justified and true.
Justified belief
What would be a justified belief?
Well there might be differences of opinion here. Some might suggest that Jeff's beliefs are only justified if they are based on well interpreted empirical evidence. He has been down on floors getting his hands dirty investigating them. Alternatively, others might feel the main source of knowledge is rational reflection – so Jeff needs to have undertaken some careful and valid philosophical investigation into floors to have worthwhile knowledge. These two sources are sometimes seen as opposed – empiricists versus rationalists.
For a scientist, knowledge is based on an ongoing enquiry where rational thought and reflection, and empirical investigation, are employed iteratively. (That is, science uses a bit of both, tending to shift between the two to build up understanding over time.)
Read about scientific knowledge
Another source of knowledge might be authority. Perhaps Jeff has been to 'floor' college and been taught by acknowledged experts? Perhaps he has carefully studied the top texts on floors and regularly tops up his expertise on professional development courses.
The founders of the Royal Society of London suggested that scientists should take no one's word for things, but rather test things out for themselves. Whilst that is a fine sentiment, it is totally impractical today, as science would never progress if all novice scientists were expected to check for themselves every piece of theory they might apply in their work. They would unlikely ever get to do anything new before retirement. Rather they can accept, on authority, that hydrogen is an element, that current in a metal wire is due to a flow of electrons, that all frogs have hearts 1, and so on. They do this on authority of their teachers and the texts they read.
Of course, this raises the problem of who can be taken as an authority.
Science is a community activity which largely works by consensus but allowing enough room for dissent to hopefully ensure it does not become dogmatic. (Sometimes it fails, but this combination of authority, logical thought, and empirical investigation is surely the best way anyone has yet come up with of finding out about the natural world.)
What is a true belief?
There is of course a distinction between a truly held belief, and a true belief. Some people truly believe they have been abducted by aliens, who having come all this way, have nothing better to do than borrow people overnight and probe them a little before returning them to their beds. No matter how much these people genuinely believe this to have happened it is not a true belief unless they really have been molested by aliens that have the technology to travel galactic distances, abduct people from inside their homes without raising the alarm, and medically examine humans without leaving any traces; but sadly have not yet found a way to do this without sufficiently anaesthetising their victims to ensure they are unaware.
Perhaps this is actually the case. If this truly happens, then someone who genuinely strongly believes they have been a victim, and has sufficient grounds to be justified in that belief (whatever we might decide would count here), can be said to have true knowledge.
But only if we know this is indeed what happened. But there is a clear danger of a vicious circle in making that judgement. We can only decide someone has genuine knowledge because we have genuine knowledge and their beliefs match our knowledge. But we can only be said to have such knowledge, if we have a true, justified, belief, which surely requires someone else to check our beliefs are true. And that some has to be someone who knows the truth – i.e., can be judged to have the knowledge.
One this basis only God or someone arrogant enough to think themselves similarly incapable of error can make this call.
Do scientists have knowledge?
It would seem that if we apply the notion of true, justified, belief strictly, then we can never claim to have knowledge. Then 'human knowledge' would only exist as some kind of idea, like an ideal gas or a canonical concept (Taber, 2019), which is a useful referent to compare things against, but not something we can ever expect to have.
That would be a reasonable way to use the word – but in actual public discourse knowledge is talked about as if it is obtainable – so experts are said to have expert knowledge (not just expert beliefs or expert opinions).
It is widely accepted that scientific knowledge claims are NOT claiming knowledge in that ideal sense – not absolute, certain knowledge, but provisional knowledge: knowledge as our best current understanding, considered to be a sufficient basis for action in the world, but always open to review.
As one philosopher of science noted:
"Assume somebody points out that certainty is an essential part of knowledge in the sense that the meaning of the word 'knowledge' contains the idea of certainty. The answer is very simple: we have decided against the idea of certainty … We have thereby also decided against knowledge in the sense alluded to. If certainty is part of knowledge, then we simply do not want to know in this sense."
Paul Feyerabend
So, scientific knowledge is evidence-based, rationally argued, and developed within an authoritative background tradition, and so considered to be reliable and trustworthy – but is strictly something other than true, justified, belief. Indeed, arguably we should not consider scientific knowledge as beliefs: scientists commit to certain ideas as well supported and worthy of provisional assent – but not as beliefs.
Read about the nature of scientific knowledge
In science education, the teacher's job is not to persuade students to believe in scientific ideas (whether seemingly controversial, such as natural selection, the big bang, or well-established such as combustion being usually a reaction with oxygen), but rather to understand why scientists have adopted certain ideas as provisional knowledge whilst remaining critical and open to new evidence or interpretations (Taber, 2017).
Assessing knowledge
If we accept that people do have knowledge (an ontological judgement), this does not imply we can easily investigate it (an epistemological matter). Science teachers assess students' science knowledge all the time, so in practice we assume that (i) in principle we can identify/examine someone else's knowledge, and that (ii) we can do this in a sufficiently valid and reliable way that we think it is morally acceptable for teachers and examiners to score and grade learners for their knowledge.
I am not disputing that is so in practice, BUT undertaking research into learners' science knowledge makes us very aware of the limitations to processes of eliciting and assessing other people's knowledge. We need probes (e.g., questions) that are understood as intended, participants motivated to reveal their thinking, and the ability to interpret responses. I am not suggesting this is impossible, but, as with any research to measure something, it relies on valid and well-calibrated instrumentation, appropriate background theory, and is always subject to limits on precision.
This is disguised sometimes in the kind of 'shorthand' used in many research reports which can include bland summary statements (of the kind '76% of the participants had a good understanding of acidity, but 12% held a common misconception') that tends to sound more absolute and precise than is possible, and which often clump together considerable diversity of knowledge and understanding within a single class such as 'misconception' or 'sound knowledge' (Taber, 2013).
In my doctoral research I studied students' understanding of chemical bonding in depth, and this reinforced the idea that each of us has unique and idiosyncratic knowledge. I found much commonality in student thinking and indeed something of a 'common' alternative conceptual framework , yet when one investigates thinking in depth (rather than just using a questionnaire with objective questions) it become clear that even in the same class, no two students have precisely the same knowledge of a topic. That certainly applied to 'chemical bonding' and I would expect it applies equally as well to other domains – including floors!
Can knowledge be quantified?
But given that everyone's knowledge is unique, with its own nuances, gaps, and deviations from some ideal canonical account, does it even make sense to try to quantity knowledge? When we appreciate some of the qualitative variation between individual epistemic agents (i.e., 'knowers') it starts to look rather questionable whether it makes sense to be satisfied with any quantitative score meant to reflect relative levels of knowledge (like test scores). This makes a claim like "no one knows more about floors than Jeff" seem, at best, a little simplistic.
A barely plausible coincidence
Yet, at the end of the first commercial break, the same sponsor made another claim:
"no one knows more about floors than Roger"
Claim made by 'Careptright'
As of itself, this claim seems just as dubious as the claim about Jeff.
But it is when they are taken together, that they really stretch credibility. It we treat these claims at face value:
- No one knows more about floors than Roger. So, this includes Jeff. Jeff does not know more about floors than Roger.
- Yet we are also told no one knows more about floors than Jeff. And this must include Roger. So, Roger does not know more about floors than Jeff.
- Clearly the only way to accept both these claims is to assume that Roger and Jeff know exactly as much about floors as each other.
Roger's domain knowledge ≥ Jeff's domain knowledge
and
Jeff's domain knowledge ≥ Roger's domain knowledge
so
Roger's domain knowledge = Jeff's domain knowledge
We have to assume here then that it makes sense to quantify domain knowledge (in the floors domain at least) and that sufficiently accurate measurements of domain knowledge can be made to confidently conclude that Roger and Jeff know precisely equal amounts.
We are not told they know everything about floors. Perhaps each only knows 10% of what it is possible to know about floors, in which case we might wonder to what extent their knowledge overlaps or is complementary. If Jeff knows 10% of all that could be known about floors, and Roger also knows 10% of all that could be known about floors, then between them they must know between 10-20% of the total domain knowledge.
Stretching credulity further
I had barely had time to ponder all this when the next commercial break arrives with a new claim:
"no one knows more about floors than Donald"
Claim made by 'Careptright'
So, an even more unlikely scenario is presented.
- No one knows more about floors than Donald. So, this includes Jeff. Jeff does not know more about floors than Donald.
- And this also includes Roger. Roger does not know more about floors than Donald.
- Yet we are also told no one knows more about floors than Jefff. And this must include Donald. So, Donald does not know more about floors than Jeff.
- And no one knows more about floors than Roger. So, this also includes Donald. Donald does not know more about floors than Roger.
Clearly the only way to accept all these claims is to assume that Donald and Roger and Jeff know exactly as much about floors as each other.
Donald's domain knowledge ≥ Jeff's domain knowledge = Roger's domain knowledge
and
Jefff's domain knowledge = Roger's domain knowledge ≥ Donald's domain knowledge
so
Donald's domain knowledge = Roger's domain knowledge = Jeff's domain knowledge
How likely is it that there are three people in the world who equally know so much about floors that no one else in the world knows more about floors than them?
I guess this deepens on just how complex a domain floor knowledge is. In many distributions that occur 'naturally' there is something of a bell-curve where a lot of people clump around the middle of the distribution, so they are fairly typical or average in that regard, and just a few people fall at the extremes.
The more possible states there are in the distribution, the more likely that the highest occupied state will have relatively low occupancy. Yet if there are only a small number of states, there will be less options and higher frequencies in specific states. That is, there will be what is known as a 'ceiling' effect. A ceiling effect occurs in tests if the questions are too easy allowing many candidates to obtain the maximum score. Such a test does not discriminate well between the most capable and less capable candidates.
There are billions of people on earth, so perhaps it is not so surprising there might be three people tied for top rating in the area of floor domain knowledge – or even more than three – so it is not impossible that these three (Jeff, Roger, Donald) all work for the same company that have highly trained them within the domain.
Beyond belief?
But then at the end of the commercial break we are told
"no one knows more about floors than Shirley"
Claim made by 'Careptright'
So, I will not go through all the argument again as clearly we are being told that no one knows more about floors than Shirley, but there are other people (Jeff, Roger, Donald) who also are not exceeded in their domain knowledge, so we conclude that there are at least four people in the world who have equal domain knowledge to each other, whilst not being exceeded in domain knowledge by anyone else.
Yet there are further commercial breaks, and further bold claims:
"no one knows more about floors than Michael"
Claim made by 'Careptright'
and
"no one knows more about floors than Johnny"
Claim made by 'Careptright'
and
"no one knows more about floors than Kate"
Claim made by 'Careptright'
and
"no one knows more about floors than Tess"
Claim made by 'Careptright'
and
"no one knows more about floors than Denise"
Claim made by 'Careptright'
So, it seems that there are at least nine people working for this one organisation who are not exceeded in their domain knowledge by anybody else in the world. So, in the domain of floors
Jeff's domain knowledge = Roger's domain knowledge = Donald's domain knowledge = Shirely's domain knowledge = Michael's domain knowledge = Johnny's domain knowledge = Kate's domain knowledge = Tess's domain knowledge = Denise's domain knowledge
And no one knows more about floors than any of them.
This seems pretty unlikely – even fantastic, perhaps.
There are several ways to interpret this.
- It is an advert meant to persuade simple people, so impressive claims are made to deceive them given that the advertisers assume their audience are too stupid to think critically and so appreciate they are being lied to.
- Alternatively, perhaps as an advert, this is knowingly making nonsense claims that the audience are meant to see through, knowing that the point of the advert is to plant an idea and brand in the mind; and the audience are not expected or required to believe it. It is not deception, but what ASA call 'puffery' – something so obviously silly that no one is expected to take it seriously (or spend time critically analysing it for a blog). Who knows, perhaps there are no such people as Jeff and his colleagues, and their parts are played by actors!
- But there is a third possibility.
Shoot high, aim low
There is one way that we might reasonably expect that a large number of people might be said to have such domain knowledge that it is not exceeded by any one else (and therefore have equal degrees of domain knowledge).
Consider a domain that is extremely simple and restricted.
We might imagine a knowledge domain with a very small number of possible knowledge elements. We can certainly imagine an artificial case – such as regarding one of those manufactured concepts used in some psychological research into concept formation.
For example consider the knowledge domain of recirgres
An example of a recirgre
Let us consider that a recirgre is a red circle on a green background. Perhaps there are only three things to be known about recirgres:
- they are circular
- they are red
- the are found on green backgrounds
There are only eight basic knowledge states here, offering four quantitative levels of domain knowledge (making a ceiling effect likely):
- knowing none of the three elements
- knowing only one of the elements (three options)
- knowing two of the three elements (three options)
- knowing all three elements
The researcher in me would point out that even here there is more potential diversity in knowledge (e.g., one person may know only that recirgre are red; another may know recirgres are red and believe they are square; another may may know recirgres are red and think they may be any oval shape including circles…but here all such options count as having one element of domain knowledge; then again, perhaps someone thinks recirgres are red circles on green backgrounds and that putting one on a North facing wall in every room of a house will bring good luck; and someone else that recirgres are red circles on green backgrounds, but they used to be princesses before they were turned into recirgres by evil stepmothers – but in either case that's just the same maximum score of 3/3 in this assessment).
If we examined the knowledge of a large number of people that had studied the topic of recirgres we might find many top scorers of whom we could say "no one knows more about recrigres than…"- because of the low ceiling in this limited knowledge domain.
So perhaps the same is true of floors. Perhaps there's not actually that much that can be known about floors, so quite a few people who work in floors and are trained up in the trade know at least as much as anyone else does. It does not take long for someone who works in floors to hit the ceiling.
Perhaps. I'd rather think that than that my viewing is being interrupted by a successions of claims that I am simply meant to dismiss as fictions that are acceptable because they are "obvious exaggerations" ("puffery") and "claims that the average consumer who sees the advertisement is unlikely to take literally" because consumers know that claims made by advertisers like 'Carpetright' are not supposed to be truthful and taken seriously. So, perhaps the advertisers do not think we are stupid, but rather that we are clever enough to know their claims are not meant to have any genuine substantive content.
But if that is so, they are not trying to deceive us, but rather to manipulate us more insidiously.
Work cited:
- Taber, K. S. (2013). Modelling Learners and Learning in Science Education: Developing representations of concepts, conceptual structure and conceptual change to inform teaching and research. Dordrecht: Springer.
- Taber, K. S. (2017). Knowledge, beliefs and pedagogy: how the nature of science should inform the aims of science education (and not just when teaching evolution). Cultural Studies of Science Education, 12(1), 81-91. doi:10.1007/s11422-016-9750-8 [Open access]
- Taber, K. S. (2019). The Nature of the Chemical Concept: Constructing chemical knowledge in teaching and learning. Cambridge: Royal Society of Chemistry.
Note:
1 Perhaps these examples seem like 'facts' rather than theories, but they can only be facts within a given theoretical context. So, to call hydrogen an element one needs a theoretical perspective on what an element is.
I've used the frog heart example to illustrate how scientific knowledge is not just as matter of being able to generalise from having examined some examples (induction):
"We might imagine a natural scientist, a logician, and a sceptical philosopher, visiting the local pond. The scientist might proclaim, "see that frog there, if we were to dissect the poor creature, we would find it has a heart". The logician might suggest that the scientist cannot be certain of this as she is basing her claim on an inductive process that is logically insecure. Certainly, every frog that has ever been examined sufficiently to determine its internal structure has been found to have a heart, but given that many frogs, indeed the vast majority, have never been specifically examined in this regard, it is not possible to know for certain that such a generalisation is valid. (The sceptic, is unable to arbitrate as he simply refuses to acknowledge that he knows there is a frog present, or indeed that he can be sure he is out walking with colleagues who are discussing one, rather than perhaps simply dreaming about the whole episode.)
The use of induction here, assumed by the logician to be the basis of the scientist's claim, might take the form:
1. a great many frogs have been examined, at different times and places, in sufficient detail to know if they have hearts;
2. each and everyone of these frogs,without exception, has been found to have a heart;
3. therefore, we can assume all frogs have hearts;
4. this, in front of us, is a frog,
5. therefore, it has a heart.
A key question then seems to be: under what circumstances can it be assumed that a property measured for some instance or specimen (something conceptualised as a member of some type or class) also applies to all other instances of the same type? …
The response is that we are not actually here using a process of induction that assumes if we look at sufficient examples we can make a generalisation across the class: rather we are using theoretical considerations. We assume that 'frog' is a kind or type such that as part of its essence (related to what makes it a frog) it necessarily has a heart. …
Frogs are a type of animal that is part of a larger group that share the same kind of circulatory system based on blood vessels and a heart to pump the blood around. We do not have to test every frog, because this type of anatomy and physiology is necessary (essential) to being a frog: it is part of the essence, the very nature, of frogness. So, the hypothetical scientist was arguing that the frog is a natural kind: a particular type of thing that exists in nature and that has certain necessary aspects, such that as long as we know we have a frog, we know these aspects will be found. So, the logical chain is something like:
1. frogs are recognised as making up a natural kind;
2. one of the necessary features of frogs is the presence of a heart;
3. therefore, we can assume all frogs have hearts;
4. this, in front of us, is a frog;
5. therefore, it has a heart.
This is not generalisation by traditional induction, but generalisation by deduction from theoretical considerations.
Taber, 2019, pp.116-119
The point is NOT that there has not been a generalisation from examining other specimens of frog, but rather that this by itself is insufficient unless located in a particular theoretical perspective.
