A paper presented to EU-translators in Brussels in 2007.
Edited by Antero Johansson
Introduction
This topic presupposes that I acquaint you with the view of language that I find scientifically well founded. To put it plainly, the experimental evidence has shown that a language consists of two interrelated parts: first, the mental space and the mental models of an individual are the most important part of language. The mental models contain the knowledge an individual has. Secondly, the language of an individual has words. They fulfil an auxiliary function: the words are the addresses that the brain uses to find the places where it has stored its mental models. It is also practical to know that scientific work has showed that the brain does not make associations between verbal entities. Instead, the brain uses addresses. I have written about this in a book (Strømnes, 2006).
Already in the early 1980’s scientists working with cognitive phenomena were aware that language has two kinds of neural entities. The view did not influence mainstream linguistics and philosophy, because at that time no one could refute the arguments used in these sciences.
Memory for words can be greatly improved
In the 1950’s it was new to American experimental psychologists that some special, mental procedures can improve memory for words and text vastly: this happens when the experimental subjects used mental images. This was new and partly shocking, because the indigenous American psychological doctrine, behaviourism, had proclaimed since around the year 1915 that there are no mental images (Watson, 1914). Behaviourist researchers tried to understand language and thinking as learned, uncomplicated reflexes.
Wallace, Turner and Perkins
Wallace, Turner and Perkins (1957) made lists of pairs of English words, which they showed one at a time to an experimental subject. The subject was told to make a mental image from the first word and another from the second word, and then to combine them. For example, if a pair was Run – Lion, then an experimental subject might make a combined image of a lion running over African plains. After the learning trial followed a recall trial. Then the subject saw the first word in each pair and wrote down the second word. Then a startling thing happened: the participants remembered all the 25 response words flawlessly.
Wallace, Turner and Perkins abandoned their quest for a limit of the effect when the list length reached 700 pairs. Then the percentage of correct answers dropped to 95%, which means that the experimental subjects recalled on average 665 response words (Fig 1)[1].
Figure 1. Results from uncommonly long lists: per cent correct answers in paired associates learning compared to typical results with conventional procedures. The figure is based on results from Wallace et al., 1957.
Kolers
In 1966 Kolers showed that a language is not a system consisting only of words. He published a study in which he used translated items in long free-recall lists. He showed that words make up one part of language, of which there are at least two. Kolers used French and English words and bilingual subjects. He found that the probability of recall approximately doubled when translated pairs were used, for instance horse/cheval (Fig 2).
Figure 2. Proportion of items recalled as a function of their frequency of occurrence in mixed and separate lists. T = translated, NT = non-translated items. Reprinted from Kolers, 1966.
J. S. Sachs
In 1967 Jacqueline Strunk Sachs published results showing that sentences are transformed into entities which are not words. She let subjects hear short tales and tested for recollection of sentences immediately after a target sentence or approximately 27 or 46 seconds after the target sentence. There were four kinds of test sentences. Three of them preserved the meaning of the target sentence. One test sentence had a different meaning.
Figure 3. Recognition of sentences that preserved or did not preserve the meaning of the original sentence. From Sachs, 1967.
Figure 3 shows that before 27 sec had passed, the subjects could not decide whether they had heard exactly this sentence before in the tale. The uppermost curve shows, however, that the meaning of the sentence was preserved when the wording of the sentence was extinguished: the subjects who heard a false test sentence were quite good at deciding they had not heard it before in the tale.
Kintsch
The influential opinion of linguists, philosophers and mathematicians has been that the brain uses words for creating propositions. Propositions were envisaged as short, special sentences made from ordinary sentences, and they are the building units of meaning. The rules for making these special sentences were found when philosophers introspected. Kintsch (1977, p.324) imported those viewpoints about meaning into experimental psychology. He wrote:
“A text is a sequence of connected sentences…
the meaning of a text… consists of a series of propositions;
propositions are composed of concepts.
Each proposition consists of one relational term and one or more arguments.”
Table 1 shows a couple of sentences with their corresponding propositions, as Kintsch gave them.
Table 1. Two sentences and the corresponding propositions. From Kintsch, 1977.
Bransford and Johnson
Bransford and Johnson (1972) showed that one must have a mental model to understand a text. They first wrote a very foggy text. People who read it rated it as very incomprehensible and recalled very little. The text is reproduced below.
Bransford and Johnson found a wondrous method for ridding the tale of its fog and making it meaningful: they showed the following picture (Fig 6). The subjects who saw this picture immediately before the obscure text found it meaningful and easy to grasp. They recalled the text well. Showing an incomplete picture did not help the subjects.
The plain explanation for this finding is that the meaning of a text is a coherent mental image. First, the brain used the words in the foggy tale as addresses to retrieve mental models. However, the parts could not be made into a coherent model because the text lacked some sentences. The picture which the experimenters showed supplied the subjects with the lacking model parts of the text.
Figure 4. The subjects who saw this appropriate model before the text understood and remembered it in the experiment. From Bransford and Johnson, 1972.
This experiment also disproves propositional accounts of language. That comes about because this theory says that images or pictures, for example, on the retina of the eye, cannot be transformed into propositions. However, a text, also a nebulous one, is transformed into propositions.
According to propositional theory showing people a picture cannot possibly aid them in understanding a text. Nevertheless, because this experiment showed that showing people a picture makes it possible for them to understand a text, propositional theories of language became disproved by that.
Strømnes 1964
In 1964 I did an experiment showing that memorising unknown words became very easy when the experimental subjects gazed at common things simultaneously with hearing the words. They learned the list at twice the speed of the subjects who did not gaze at things. They found it easy to recall the list backwards, for they either looked at the things or at their mental images of the things. This served to bring the words to mind. Figure 5 shows the difference in difficulty.
Figure 5. Memorising unknown words while gazing or not gazing at common things. From Strømnes, 1964.
The use of mental images or pictures improves performance: increasing the work load of the subject makes the task considerably easier. Additionally, the use of mental images and pictures in linguistic tasks diminishes the differences between the subjects. In the experiment in question, the subjects who looked at things differed little, while the subjects who did not look at things differed greatly from each other.
Moeser and Bregman
Two articles by Moeser and Bregman (1972; 1973) presented the experiments that best have revealed how people learn language. Shannon Dawn Moeser made a small experimental language. She also made up the miniature world that the experimental language could speak about. In the miniature world there were coloured rectangular and non-rectangular shapes. The shapes could have different orientation, different borders and spatial relations.
Moeser let one group of the subjects in the study learn the language without seeing what the sentences meant in the miniature world. The other group saw a sentence and simultaneously what it meant. Figure 6 shows an example of this. Translated into English the sentence means approximately “on the left there is a green rectangle with a dotted border and on the right, there is a square with a filled border. The square has the same height as the rectangle.” This is also an example of how a short, unequivocal sentence in one language becomes a long, unwieldy one when translated into another language.
Figure 6. Examples of designs when learning a miniature language. Adapted from Moeser and Bregman, 1972.
Figure 7. The subjects learned the miniature language well when they saw the model geometry of the language. The other subjects remained at chance level (a score of 52,5 points). The figure is based on table 2 in Moeser and Bregman, 1973.
However, even with 3200 repetitions the subjects who did not see what the sentences were about, did not learn to understand the language at all. The subjects who had the extra load of perceiving the spatial structures of the sentences started to learn the language after about 400 presentations. Soon they had learned the language perfectly. Figure 7 shows this.
Thus, Moeser showed that a language cannot he learned by only studying its sentences.
Paivio
The Canadian experimental psychologist Allan Paivio did a multitude of well-designed experiments about words and mental images during the 1960’s and 1970’s. He published his first results in a book appearing in 1971. Paivio found that words differ in their ability to activate images in the mind. He showed that this ability can be measured and scaled as a value of imagery. Paivio showed the imagery characteristic of words to be the variable that most strongly influences the outcome of experiments on verbal behaviour. In his famous book from 1971 he expressed the hypothesis that the child learns to understand sentences when perceiving the spatial order that belongs to the language.
Strømnes on language structure and language learning
The experiments I have spoken about contain forceful aid to practical work with language. Already from the late 1960’s I had to make practical use of the insights about language I have described here. The reason was that I wanted to continue my work towards the doctorate in Finland, in a position as private reader. Then the rules said that a private reader had to have a perfect command of the Finnish language, and this regulation was no idle intimidation. Here there was no question of tempering justice with mercy.
Sad to say, the painful truth soon dawned upon me: evidently, I could not learn Finnish well. For, I could not learn to understand the meaning of the Finnish cases in the way the Finns do. Consequently, in the laboratory we started a search for the spatial structure of the Finnish cases in. I devised an experimental situation with which we could measure precisely whether we had succeeded. Figures 8a and 8b show some examples of the pictures for Finnish cases and for Swedish prepositions (Strømnes, 1974a).
Figure 8a.Three models that immediately called corresponding case addresses to mind for Finnish-speaking subjects in the experiments by Strømnes. From Strømnes, 1974a.
The pictures of the Finnish genitive and the partitive are approximately correct. Therefore, they do not have full functionality. When we had perfected all the pictures, we did not find a publishing outlet.
Figure 8b.Three models that immediately called corresponding preposition addresses to mind for Swedish-speaking subjects. From Strømnes, 1974a.
Figure 8a shows you why I could not understand how to use the partitive and the genitive correctly. The Finnish case structures show relations between bordered entities staying in a mental space with two spatial dimensions and a time dimension. This means that the cases contain movements that last for different amounts of time.
The Swedish preposition structures show a relation between a trajectory and a comparison entity. The structures lack a time dimension and have three spatial dimensions. The preposition structures in Norwegian resemble the structures in Swedish, and in other Indo-European languages. Starting from my unconscious knowledge of Norwegian, I could not guess the Finnish structures. However, the long way through the laboratory solved the problem.
When we had found and perfected the spatial structure of a case, I could use it. For example, whenever the Swedish sentence had an object, I inspected the mental image that the Swedish sentence created in my mind. After that I compared the image to our pictures of the partitive and the genitive. So, for the translation I chose the case that closest resembled the object part of the Swedish mental image. This way to translate worked immediately and perfectly. My object errors vanished. The thing is that a correct grammar has no exceptions and is extremely simple to use. It shows the correct structures in mental models and nothing more.
The pictures of the Swedish preposition structures were equally effective for Finns who experienced difficulties with Swedish.
My assistant Iivonen and I showed that we could solve the syntax learning problem considered the most difficult one: deaf children with undemanding intellectual capacities earlier could not be taught to understand or to make sentences. However, we solved this problem and published a paper about it (Strømnes & Iivonen, 1985).
Finke
Many experimental psychologists did voice doubt and criticism about the reality of mental models. For example, the message of the experiments was doubted, because the subjects could have used devious schemes to improve their performance. Many studies met with such criticism. However, criticism fell silent after 1980, when Finke (1980) published a paper about the problem. His experiments showed that pattern analysing in the visual system goes on in the same way when things are imagined as when they are observed.
The arguments in favour of propositions
Nevertheless, the cognitive role of the mental models still did not become well acknowledged. Highly probably the reason was that researchers in the USA could not or maybe did not want to disprove some arguments about images and language that had originated in philosophy and mathematics. The arguments were intended to show that images or other likenesses cannot store knowledge. Additionally, they maintained that knowledge can be immaterial and purely spiritual.
In my latest book (Strømnes, 2006) I showed that the arguments denouncing mental models and positing immaterial knowledge are wrong. These arguments can be dismissed.
Thus, I have now presented a well-founded view of language for you. Now I can consider what can be said further about the theme of this lecture: Are there significant differences between the mental outlooks of languages?
Significant differences
The idea of “mentality” is absent from experimental psychology. This may he comparable to the fact that our climate exists. Still it is not a variable in the science of physics, which, however, is used to explain our climate.
Mentality can be understood to mean patterns of perception, thought and behaviour that are characteristic for a group of people. There is a tradition among ethnolinguists and anthropologists that supports the view that such patterns are caused by or bound up with the language people use. In the USA Benjamin Lee Whorf (Whorf, 1956) popularised the view of Karl von Humboldt. In the USA the view became known as the Whorfian or the Sapir-Whorf hypothesis.
We should be aware that the view researchers hold about this question affects the interests of political and financial circles, and of some traditionalist groups. For example, if language has no determining influence on thought and behaviour, then minority languages can be eradicated or overlooked without this having any deleterious consequences, and majority languages can be imposed everywhere. There is a vast language teaching industry, which thrives very well with the conception that languages everywhere communicate and determine in the same way. Conditions like these, may, of course have many deplorable effects on the scientific study of the effects of language.
In the USA the study of differences between the mental outlook of languages has, mostly, been a field where anthropologists and cultural psychologists have made studies. They have essentially concentrated on colour ideas in different languages, and they have found very little of interest. That could easily be predicted, because colour vision depends on the neurochemical constitution in the light-sensitive cells in the retina. Language can, of course, not influence these cells in important ways. John A. Lucy, who in 1997 published a review paper on Linguistic Relativity in the Annual Review of Anthropology, wrote the following.
‘The neglect of empirical work is so conspicuous that it must be regarded as one of the central characteristics of this area of research and warrants brief comment”.
He then went on to comment on the fears, preconceptions, “prejudices and passions” that have discouraged research in the area.
Recently, two German researchers, Bornkessel-Schlesewsky and Schlesewsky (2007) have showed that some areas in the brain react dissimilarly to word order and to use of cases for people who speak different languages. Their findings show that there are cerebral differences between languages. It is paradoxical, when Bornkessel-Schlesewsky and Schlesewsky try to resolve the new observations by entirely traditional linguistic means. For example, they assume that language can he described by two-dimensional tree diagrams.
Tools
When thinking about similarities and differences between languages, one must reflect on the relation between a tool and the organism using it. A language has the character of a tool because it can he used to handle tasks. Here (Fig 9) I show you three tools, a saw, a chisel and a plane. I take the saw and the chisel one after the other in my right hand, and I also show you how I hold the plane with both hands. Thus, you saw that a tool has a part that meets or is embedded in a part or structure of the body. The contact area of a tool must be similar with the contact areas of other tools that integrate with the same part of the body. The figure 11 shows this. As you see, the tasks you can do with tools with similar contact areas can be quite different from each other.
Figure 9. Examples of different tools.
For example, words can sound very different across languages, but words still must stay within the boundaries of what the speech organ can do. Also, the mental space and its models must be of the kinds that the brain can hold.
Our comparative research with Finnish and Swedish
Returning then to our comparative research with Finnish and Swedish, we may recall the models we had found of the Finnish cases and the Swedish prepositions (Strømnes, 1974a). We can take a short glance at them again (Figs 8 a-b).
The Finnish system depicts the relations between borders. In the Swedish system, the concern is with relations between movement trajectories and an object of comparison.
Judging incomplete sentences
We did an experiment that involved judging incomplete Finnish and Swedish sentences (Strømnes, 1974b). They were exact translations of each other, for instance the Finnish sentence “Isäntä avasi paketin kellarissa” and the Swedish sentence “Husbonden öppnade paketet i källaren”. The two sentences mean “The master of the house opened the parcel in the cellar”.
The following four figures (10 a-d) show that sentences which lacked a verb were judged as more Finnish and less Swedish. Consequently, this simple experiment disproves theories holding that the mental function of a verb is the same in all languages.
Figure 10b. Judgement of the Finnishness and the Swedishness of the deficient sentences in the object set. Percentages of total scores are shown. From Strømnes, 1974b.
Figure 10c. Mean scores for the sentences without a verb in the adverbial set. From Strømnes, 1974b.
Figure 10d. Mean scores for the sentences without a verbin the object set. From Strømnes, 1974b.
The same sentence material was used for two other tasks (Fig 11 a-b). First, Finnish and Swedish subjects were asked what deficient sentences a detective short of money would choose to send a telegram to his headquarters. The Finnish subjects let the detective drop the verb much more often than the Swedish subjects, who rather kept the verb.
In the second task the subjects were made aware that an artist could produce a drawing from the base sentence. In this task the Finnish-speaking subjects judged that the artist could do excellently without knowing the verb of the base sentence.
Figure 11a. The experimental subjects let the imagined detectives drop adverbial operators equally often. However, the Finnish one strongly preferred to delete the verb. The Swedish detective tried to keep it more often. From Strømnes, 1974b.
Figure 11b. The choices of sentences without a subject and without a verb for drawing a picture, as percentages. From Strømnes, 1974b.
Taken together, the Finnish-speaking subjects thought it equally unimportant for the detective’s headquarters and for the artist to know about the movement in the mental model. The Swedish speaking subjects judged differently. They found it more important for the artist than for the detective’s headquarters to know about the movement in question.
This outcome is extraordinarily important. It shows that one mental model geometry causes its users to consider two tasks to be equal and to require the same treatment. Another mental model geometry causes its users to judge the same tasks to be different and to require dissimilar treatment.
Discarded words in headlines
When people are short of time and space, they are prone to let go of parts of their sentences. They omit those parts that they think the receivers need the least, or most easily can reconstruct. I studied 4809 headlines from the largest Finnish and 3442 headlines from the leading Swedish newspaper in Finland (Strømnes, 1974b). The editors in both newspapers quite often let people guess the action that their headlines suggested. However, the Finnish editors did it more often, and the difference reaches a high level of statistical significance.
Ice-hockey speech
Radio reporters who speak about sporting events also often omit parts of their speech. We recorded ice-hockey matches where a Finnish-speaking and a Swedish-speaking reporter covered the same match (Strømnes, 1974b). However, they perceived the events quite differently. The Finnish-speaking reporters let go of more verbs and spoke more about topological events. For example, they spoke about groupings of people. The Swedish-speaking reporters tended to keep the verb and to speak about vectoral events. For example, they spoke about the movements of the puck.
The bilingual research assistants who coded the tapes said it was a weird experience to listen simultaneously to both reporters of a match. It was weird because the assistants knew the reporters had been present at the same match, but the images they conveyed from it were strikingly dissimilar.
Comparing Finnish and Swedish moving pictures from the same texts
TV-films produced in Finland are strikingly unpopular in the other Nordic countries. We studied the cause for this (Strømnes, Johansson, and Hiltunen, 1982). For our material we chose classical plays that had been filmed by a Finnish team in Helsinki, and by either a Swedish team in Stockholm or a Norwegian-speaking team in Oslo. Then, for each play there was one text and two realisations.
When we turned off the sound tracks in the laboratory, it was hard to recognise that the two films in a pair were about the same topic and from the same manuscript. The reason was that the way the pictures were made in the two versions was strikingly different. In the Finnish versions there were more close shots, which also lasted longer than they did in the Scandinavian versions.
Figure 12a. The mean length of all movements, measured in steps, in the Finnish and Scandinavian versions of Tartuffe, A Doll’s House, and Uncle Vanya. From Strømnes et al., 1982.
In the Scandinavian versions there were considerably more long movements within shots, and there were many movements continuing in another shot size. This means that the Scandinavian versions concentrated on the interrelated movement patterns of the characters, and these patterns were shown in three-dimensional space.
The figure 12a first presents the mean length of all movements in the six films from three texts. Figure 12b shows the amount of interrupted movement in the pictures. On the left in figure 13 is shown how many movements had a length of five or more steps. On the right is presented the number of steps that were in depth and in plane.
Figure 12b. The combined amount of interrupted movements in each of the Finnish and Scandinavian films, presented as percentages of the total amount of change of location. From Strømnes et al., 1982.
Figure 13. Movement with length five steps or more in the Finnish and Scandinavian films. The total sum of steps is shown, and the proportion used for movements in depth, and in plane. From Strømnes et al., 1982.
These charts mean there is considerably more study of human relationships in the Finnish versions than in the Scandinavian ones. The Scandinavian movement-centred realisations showed much less interest in the study of human emotionality.
It is important to notice that the findings from the film comparisons parallel the earlier case and prepositions study: the spatial structures we found then, now were discovered to be the structural base on which the scenes in the studied films were built. This is a demonstration of a close correlation between the spatial structure of mental models and behaviour.
Occupational accidents
It is possible that the mental model structure may govern behaviour generally, and so can have serious societal consequences. This might be the case in Finland, where the country has imported a technology that Indo-European speaking people created. The technology is made for and by people who are trained to perceive movement in three-dimensional space. In its turn this would mean that people who are trained to observe the relations between borders may come to overlook potential sources of danger in the technology. Therefore, it could be predicted that work accidents would he more frequent among Finnish speaking employees and workers than among otherwise perfectly comparable Swedish speaking employees and workers. In Finland can be found groups of gainfully employed people who only differ as for language. Their wages, education, and all other societal conditions are the same.
Figure 14a. Occupational accidents per 1000 workers in three sources: The Farmers’ Pension Institute, 1991; The National Board of Labour Protection, 1990; Statistics of Finland, 1988. The counties were: Uusimaa, Turku and Pori, and Vaasa. From Strømnes, 2006.
However, many years elapsed before the Institute of Occupational Health in Helsinki lent ears to this hypothesis. When they did, they found that the hypothesis was perfectly correct. All sources of information about comparable Finnish and Swedish speaking groups showed that the Finnish speaking employees and workers had more occupational accidents than the Swedish speaking ones (Salminen et al., 1996; Johansson & Salminen, 1999; Salminen and Johansson, 2000). Figures 14 a-c shows this.
Figure 14b. Accident frequencies of Finnish and Swedish speaking workers (N=33684; N=8359) in the county of Vaasa according to records of the National Board of Labour Protection in 1990. From Strømnes, 2006.
Figure 14c.The accident frequencies of Finnish speaking (n=9202) and Swedish speaking (n=485) blue-collar, white-collar, and self-employed persons in a random sample from the Finnish population in 1988. From Strømnes, 2006.
The important message of this investigation is that only the difference between the Ural-Altaic and the Indo-European mental model system can explain the findings. For, the two groups have the same culture and the same geographical environment.
Conclusions
Apparently further research about the mental geometries of languages could be widely extended. Then we might come to have new, efficient tools for making wise decisions.
I thank you for coming and listening to me.
Remarks
From Wallace, Turner and Perkins
[ For example, the subjects who saw 700 pairs of words had to remember 1400 single words, and in addition, the 700 image combinations of the list.]
From Kolers
[ Translation 1 (for instance, horse) and translation 2 (for instance, cheval) have different physical forms. Then, the neural counterparts of these words are different, too. Still, presenting one of them improves memory for the other one. Therefore, the translations 1 and 2 must lead to and start a third entity, which they have in common. This shared entity is a mentally more important part of language, because it receives the main increase in memory strength. Thus, this study of Kolers suddenly showed that the age-old view of language was sorely deficient: a language is not a system consisting only of words and their neural counterparts.]
From Sachs
[ One might, of course, assume that the meanings the brain builds from sentences are mental models, since the experiment of Wallace, Turner and Perkins showed that the brain uses words as addresses to retrieve mental images. (By the way, the modern term mental model includes both conscious and unconscious mental images.)]
From Bransford and Johnson
[ In psychology “perception” means acquainting oneself with an entity through the senses. When words are addresses for mental models, then language learning necessarily includes perception of the structure of the environment. In addition, the learner must perceive a sound that can serve as an address for the percept.]
Strømnes on language
[ Consequently, the sentences I said and wrote were often distressingly wrong. I refused to think that this inability of mine was my own fault. Small Finnish children could, at that time, speak Finnish quite well, while I who constructed experiments for a doctorate, could not learn the language.]
From significant differences
[ In Europe, Karl von Humboldt (1767-1835) was the first well-known exponent of this view. The anthropologist Franz Boas (1858-1942), who was born and educated in Germany, introduced the view of Humboldt in the USA. His student Edward Sapir (1884-1939) also taught this theory. Boas and Sapir had studied American Indian languages, and so did Sapir’s student Benjamin Lee Whorf (1897-1941).]
Judging incomplete sentences
[ The subjects were native speakers of either Finnish or Swedish from high schools selected from areas where there was very little likelihood that they would use any language other than their mother tongue in daily communication. In the first task (Figs 10 a-d) there were 39 Finnish-speaking and 42 Swedish-speaking subjects, in the second (Fig 11a) 14 and 20, and in the third (Fig 11 b) 20 and 20, respectively.
Two sets of 10 base sentences in Finnish and Swedish were made. In the adverbial set the structure in both languages was, subject, verb, adverbial, and adverbial (for example: The plane flew from Vaasa to Porvoo). From each base sentence a subset of sentences was derived by deleting, one item at a time, the subject, the first adverbial operator, and the second adverbial operator. These sentences were all incomplete, one of them lacking a noun in the nominative (the subject) and one of them lacking the active verb.]
Occupational accidents
[ I had worked as a research scientist in Finland for almost three decades before the investigation about occupational accidents was done. Therefore, I, of course, knew before the investigation started what its outcome would be. The frequency of occupational accidents in society is a factor that very decisively influences the prosperity and well-being of nations. Because we knew so much about work accidents, obviously our knowledge could save impressive amounts of pain and money if the knowledge was put to use.]
======================================================
[1] Editor’s remark: Senter et al. (1982) did a quasi-replication of the Wallace et al. (1957) experiment. One of their two subjects exhibited 75.8 % correct recall of 348 PA response members and 87.5 % of 154 pairs of concrete nouns while under instructions to use a mental imagery learning strategy. The other subject only remembered 25.5% correct which was worse than the controls exhibited on average (40.07%) after studying a list of 100 pairs printed on paper without any instruction to use imagery.
References
Bornkessel, I., Schlesewsky. M. The extended argument dependency model: a neurocognitive approach to sentence comprehension across languages. Psychological Review, 2006, 113, 787-821.
Bransford, J. D., Johnson, M. K. (1972) Contextual prerequisites for understanding: some investigations of comprehension and recall. Journal of Verbal Learning and Verbal Behavior, 11, 717-726.
Bransford, J. D., McCarrell, N. S. (1974) A sketch of a cognitive approach to comprehension: Some thoughts about understanding what it means to comprehend. In: Weimer, W.B. and Palermo, D.S. (eds.) Cognition and the symbolic processes. Hillsdale: Lawrence Erlbaum.
Finke, R. A. (1980). Levels of equivalence in imagery and perception. Psychological Review, 87, 113-132.
Huttenlocher, J. (1968) Constructing spatial images: a strategy in reasoning. Psychological Review, 75, 550-560.
Johansson, A. Salminen, S. (1999) A minority with few occupational accidents: The case of Swedish-speaking Finns. American Journal of Industrial Medicine, Supplement 1, 37-38.
Johnson-Laird, P. N. (1983). Mental models. Towards a cognitive science of language, inference and consciousness. Cambridge: Cambridge University Press.
Kintsch W., Kozminsky E. (1977) Summarizing stories after reading and listening. Journal of Educational Psychology, 69, 5, 491-499.
Kolers P. A. (1966) Interlingual facilitation of short-term memory. Journal of Verbal Learning and Verbal Behavior 5, 314-319.
Kosslyn, S. M. (1980). Image and mind. Mass.: Harvard University Press.
Lucy, J. A. (1997). Linguistic relativity. Annual Review of Anthropology, 26, 291-312.
Moeser, D. S., Bregman, A. S. (1972). The role of reference in the acquisition of an artificial language. Journal of Verbal Learning and Verbal Behavior, 11, 759-769.
Moeser, D. S., Bregman, A. S. (1973). Imagery and language acquisition. Journal of Verbal Learning and Verbal Behavior, 12, 91-98.
Paivio, A. (1971). Imagery and verbal processes. New York: Holt, Rinehart, and Winston.
Sachs J. D. S. (1967) Recognition memory for syntactic and semantic aspects of connected discourse. Perception and Psychophysics, 2, 437–442.
Salminen, S., Johansson, A., Hiltunen, E., Strømnes, F. J. (1996). Suomen- ja ruotsinkielisten työntekijöiden tapaturmasuhteet. Työ ja ihminen, 10, 125-136. (A comparison of the frequency of occupational accidents among Finnish- and Swedish-speaking workers.)
Salminen, S., Johansson, A. (2000) Occupational accidents of Finnish- and Swedish-speaking workers in Finland: A mental model view. International Journal of Occupational Safety and Ergonomics, 6, 293-306.
Senter, R. J., Richter, D. O., Wilson, S. D. et al. (1982) Wallace, Turner, and Perkins revisited Bulletin of the Psychonomic Society, 20, 178-179.
Shepard, R. N., Metzler, J. (1971). Mental rotation of three-dimensional objects. Science, 171, 701-703.
Strømnes, F.J. (1974a). No universality of cognitive structures? Two experiments with almost perfect one-trial learning of translatable operators in a Ural-Altaic and an Indo-European language. Scandinavian Journal of Psychology, 15,300-309.
Strømnes, F.J. (1974b). To be is not always to be. The hypothesis of cognitive universality in the light of studies on elliptic language behaviour. Scandinavian Journal of Psychology, 15,89-98.
Strømnes, F.J. (2006). The fall of the word and the rise of the mental model. A reinterpretation of the research on spatial cognition and language. Frankfurt am Main: Peter Lang.
Strømnes, F.J., Johansson, A., Hiltunen, E. (1982). The externalised image. A study showing differences correlating with language structure between pictorial structure in Ural-Altaic and Indo-European filmed versions of the same plays.Helsinki: The Finnish Broadcasting Corporation, Report No. 21/1982.
Strømnes, F.J., Iivonen, L. (1985). The teaching of the syntax of written language to deaf children knowing no syntax. Human Learning, 4,251-265.
Wallace, W. H., Turner, S. H., Perkins, C. C. (1957) Preliminary studies of human information storage (Signal Corps Project No. 1320c). Philadelphia: University of Pennsylvania, Institute for Cooperative Research.
Watson, J.B. (1914) Behavior: An introduction to comparative psychology. New York: Holt.
Whorf, B.L. (1956). Language, Thought, and Reality. Selected writings of Benjamin Lee Whorf. Edited and introduced by J.B. Carrol. Cambridge, Mass.: MIT Press.