On Time, Space, and Metaphor II: A Proposed Experiment

I've got it! I know how we can test conceptual metaphor theory in the domain of time. Yesterday I argued that the problem with the experiments published so far is that it's impossible to distinguish mere lexical priming (priming spatial meanings of the terms, which influences subsequent temporal reasoning) from conceptual priming due to metaphorical mappings between time and space. As I was walking across campus today, I came up with an idea. Now, this idea is in its infancy, but I bet some of you can help me with it. Of course, I don't actually plan on running this experiment, because coming up with the stimuli is going to be a lot of work (I'll explain why in a minute), but there are tons of cognitive linguists out there with nothing better to do, so if upon further reflection, and any revisions you and I can come up with, the experiment seems like it might actually help to solve the problems with the previous experiments, we can always suggest it to them.

If you recall from the post on false memories in bilinguals the other day (assuming you read it, which is a big assumption), within the Deese-Roediger-McDermott paradigm, memory accuracy tends to improve for both list words (words that were actually on the to-be-remembered list) and critical lures (words that are primed by the words on the to-be-remembered list) after the first few recognition trials. This is generally believed to be the result of the loss of conceptual information in memory, and an increased reliance on lexical information (information about the words themselves, and not their meanings).

You can probably see where I'm going with this. Conceptual metaphor theory states that our concept TIME is metaphorically structured, and receives much of its content, through its mapping to our concept SPACE. In order to provide nonlinguistic empirical evidence for this, we would need to sort the conceptual information from the influence of the lexical information shared by the words we use to talk about time and space. Here's how I think we could do this. We compile two word lists that only include words that are used to talk about time, and not space. One list would prime the concept FUTURE, and the other would prime the concept PAST. These will be the target lists. Each participant would learn one target list, along with several other word lists that prime concepts unrelated to time or space. In the memory test, participants would receive a word list that includes some words that were actually on the word lists they saw, along with the usual critical lures for the non target lists. However, in place of the usual lures for the target lists (FUTURE or PAST), they would receive the spatial terms used to talk about FUTURE and PAST, namely "in front" and "behind." Again, if TIME is metaphorically structured through its mappings to SPACE, we would predict that priming FUTURE would also prime IN FRONT, and priming PAST would also prime BEHIND. Therefore, we would expect to see false memories for the spatial words on the first couple memory trials, and over time, the rate at which participants falsely remembered having seen these words on the target lists would drop significantly. This would be a clear indication that the TIME concept includes conceptual information from the SPACE concept.

What do you think?

UPDATE: I forgot to say why it would be difficult to come up with the stimuli for the experiment. To come up with word lists for use in the DRM paradigm, you have to first show that your lists work. That means you have to show that they prime your concept in the vast majority of participants (80-90%). In this case, that's made all the more difficult by the fact that in the target lists, you can't use spatial terms, and many of the words we use to talk about the future and the past are spatial, which was the inspiration for the theory in the first place. And you need 10-12 words for each list.

More like this

False memory research has been very popular over the last several years, in part because of its connection to one of the more contentious debates in cognitive science: the recovered memories debate. I remember attending a poster session at a conference a couple years ago, and an entire room was…
I've said it before, and I'll say it again: conceptual metaphor theory sucks. Why does it suck? Well, because there's no experimental evidence for it (and plenty of evidence against it). Except, that is, in one domain: time. Specifically, the work of Lera Boroditsky, along with Dedre Gentner and…
As I've said before, the primary (if not the only) real experimental evidence for conceptual metaphor theory -- the theory that abstract concepts are structured by more concrete (i.e., closer to sensory/perceptual) concepts -- comes from one domain: time. Time, according to conceptual metaphor…
Cognitive scientists are increasingly aware of how individual differences can confound experimental results. That is, differences in group means cannot always be interpreted clearly if, for example, only some subset of individuals in each group demonstrates the effect. Consequently, even the…

Here's my attempt at brainstorming... I wasn't sure whether the words have to be in any particular form (nouns?), and I've tried to purge spatial words, but it's tricky that a lot of the words I've come-up with that aren't directly spatial are closely associated with other words that are (e.g. memory --> flashback).

weatherman (forecast is probably too spatial?)

grave digger
gone (too spacial?)
long ago
olden days

I think this is a really good idea for an experiment. One potential problem, though, is the supposed unidirectional mapping of metaphorical concepts. that is, we think of time in terms of space but not space in terms of time. one of lera boroditsky's students, dan casasanto, showed that the perception of time -- in this case how long in seconds a line was present in a display as it expanded horizontally -- was influenced by the actual length of the line in centimeters (i.e. how spatially extended the line became). the longer the line in centimeters, the longer in duration people judged the line to be present in the display. however, the reverse situation did not hold: the duration of the line in the display did not affect people's judgement of how spatially extended the line was. the idea is that the concept of time is structured via metaphorical mapping to the concept of space, but time perception in no way structures spatial perception. if your experiment failed to produce any false memories for spatial terms, it could be that it is because our time concept does not actually prime specifically spatial concepts at all (the metaphorical mapping being unidirectional from space to time). incidentally, i think casasanto's work provides some potentially strong nonlinguistic evidence for how we structure time through space, though the effects are small and i don't think the work is published yet (i know about them because i saw his thesis defense talk at MIT last fall). what do you think?

Steve, I think I remember either hearing of Casasanto's work, or seeing it presented at a conference (as a poster, I think). I'd have to look at the data to know what I really think. It definitely sounds promising.

As I was thinking about the experiment I described in this post, I was thinking about testing it in reverse, too (I originally thought four lists, FUTURE, PAST, IN FRONT, and BEHIND), but it would be exceptionally difficult to come up with the spatial primes, because just about any spatial word can be used to talk about time, whereas the reverse is not true. However, I don't think it would be a problem for the time lists. If the concept FUTURE is structured with the concept IN FRONT, then anytime you prime FUTURE, you will also prime IN FRONT. That relationship may be unidirectional (priming IN FRONT might not prime FUTURE), but if conceptual metaphor theory is true for TIME, then the time-to-space direction should work.

I found Casasanto's paper. It's under review right now, it looks like. You can read it here.

Notice the running head. I feel so uncreative.

OK, I read the Casasanto and Boroditsky paper while I was eating lunch. It's actually very interesting. All six experiments are variations of the first, so if anyone reading this wants to check out the paper, but doesn't want to wade too heavily into the methodological details, you can get away with just reading Experiment 1.

I'm torn about whether to post about it until it's at least been excepted somewhere, but I'll give a couple quick impressions here.

1.) You can't account for their data with the sort of priming effects I talked about. The data really does seem to reflect the influence of space on time, conceptually.

2.) I still don't know how metaphor fits into it, but it is what you would predict from conceptual metaphor theory. Now, I have some data of my own that can't be accounted for by priming, either, but it paints a slightly different picture. I may post about that, too.

It definitely does seem to be consistent with an embodied, and perhaps modal representation of time, though. And that's even more interesting.

I also read over the paper, and I agree with your assessments. It is especially interesting that the effects were so reliable in this non-linguistic task, and it opens the door for more of this type of psychophysical experimentation. I am also interested in your data that you mentioned in your post and what picture it paints... at this point i might as well come clean and admit my own bias: i am starting my phd in cog psych at stanford next year with lera as my advisor (and, like lera, i did my undergraduate work at northwestern under dedre gentner). personally i agree with you that much of the work on conceptual metaphor is ambiguous at best, but i think lera's group is probably the most innovative and creative of them all and seems to be gathering more and more interesting data...

hey Steve, congrats on starting on the path to becoming part of the higher-order cognition mafia (with Dedre and Doug as the co-Godparents).

I agree with you 100% that Lera's work has been the most innovative in conceptual metaphor theory. In fact, I'd say that it's miles and miles above anything any cognitive linguists have done.

i'm a bit wary of talking about my own work here, because it would reveal my secret superhero identity, but if you email me, I'll tell you a little bit about it.

Hi all,

There was a recent article that mentioned that Amayara do not think of future as in front and Past as in back, but just the reverse. The link is http://ucsdnews.ucsd.edu/newsrel/soc/backsfuture06.asp

Also, in one of my own blog entries, http://the-mouse-trap.blogspot.com/2006/06/cognitive-map-continued-its-…, I argue that Indians (using Hindi or other similar languages) have the same word 'kal' used for both tomorrow as well as yesterday, so they may be prone to seeing the time as circular. This may be counterintutive, but the mapping to space in this case may be to a space represnted by using angular geometry(r,theta) and hence the perception of time as cicrcular. This explains the recurring metaophorical and mythical themes in such civilisations like 'periodic recurring epochs' , 'recarnation" etc.

Some food for thought as to some other Metaphors that might be used for space (one of my blog entry focuses on angular geometry concpets of space in not only men, but also mice and time) and time.

A very mind-broadaning article though. Kudos to you!