Blending in Context

Summer School
Odense, Denmark
August 12-17, 2002

Organised by the Graduate School in Language and Communication, University of Southern Denmark

Brief summary and comments by Pierre Gander (pierre.gander@lucs.lu.se)

Index

Faculty
Seana Coulson (University of California at San Diego)
Margaret H Freeman (Los Angeles Valley College)
Todd Oakley (Case Western Reserve University)
Tim Rohrer (University of California at San Diego)

Participants
Line Brandt, University of California, San Diego
Natalie Cigankova, University of Latvia
Vito Evola, University of Palermo
Lise Lyng Falkenberg, University of Southern Denmark
Laura Feldt, University of Copenhagen
Maurizio Gagliano, University of Bologna
Pierre Gander, Lund University
Monica Gonzalez-Marquez, Cornell University
Anders Soegaard, University of Copenhagen
Michael Kimmel, University of Vienna
Irene Mittelberg, Cornell University
Francisco Peireira, University of Coimbra
Paul Sambre, LESSIUS Hogeschool Antwerp/University of Antwerp UFSIA
Hedvig Gyde Thomsen, Copenhagen University

Introduction

The summer school was on blending theory, a theory of conceptual integration by Gilles Fauconnier and Mark Turner (see, for instance, their book "The Way We Think") In the summer school, the aim was to show how blending theory can be applied to various contexts, such as literature and computer interfaces, as well as giving a better understanding of blending theory itself. The school included lectures as well as workshops and exercises. PhD students were encouraged to present their work.

Day by day overview and comments

August 13

Topic: Narrative context: blending in literature

Margaret Freeman presented a blending analysis of an untitled poem by Emily Dickinson ("The Luxury to apprehend...") and showed that by using blending theory, meanings before not considered by literary critics can surface. Freeman talked about poetic creation versus poetic comprehension. She said: "Blending theory is an elegant theory for creativity". She concluded that some elements of poetry are not yet handled by blending theory, such as rhythm, sound, which need to be explored.

Monica Gonzalez-Marquez presented the lyrics to a latin song with the translated title of "Her and him" for subsequent blending analysis later in the summer school.

Guest lecture by Robert E. MacLaury (University of Pennsylvania)
MacLaury gave an overview of vantage theory, which is a model of categorization. He focused on applying the theory to colour categorization in various cultures and languages, but also showed how the model can account for forms of addressing and deixis.

August 14

Topic: Blending in Interaction

Seana Coulson gave and introduction to blending theory and showed how it could be applied to an example of talk from a radio station.

Gitte Rasmussen gave a short introduction to Conversational Analysis. Anders Hougaard showed how Conversational Analysis can be extended with blending theory. Analysis was made of a telephone conversation from the Watergate affair between Colson and Hunt.

Line Brandt presented a blending theory analysis of a segment from a TV show focusing on the metaphor "digging your own grave".

Irene Mittelberg presented PhD work on cognitive models of grammar. She showed gesture data on language teachers' metaphorical gestures of grammar concepts.

August 15

Topic: Blending in Grammar

Todd Oakley presented his theory of attention for blending theory. It is an important area that is missing in blending theory, he stated. The theory, drawn together from knowledge in various fields such as psychology and neuroscience, was very sketchy and provided a heuristic and a vocabulary rather than a proper theory.

Anders Soegaard presented PhD work on nominal compounding in Mandarin Chinese.

Small group work on examples. Our group worked with an example with language use were an animal was blended with machines (or possibly toys) ("Snake is broken").

Chris Sinha gave a lecture on the concept of emergence, with background from biology and philosophy of science. The aim was to give a ground for the emergence concept in blending theory.

August 16

Topic: Computers and Interfaces

Tom Rohrer lectured about how blending analysis can be applied to computer interfaces. The participants analysed, using blending theory, an example in the form of a web commercial banner graphic. Rohrer gave some general hints for how to start making a blending analysis: 1) Brainstorm a list of lexical, visual, etc. items in the example to be analysed 2) Label some input spaces

Paul Sambre presented his PhD work on analysis of copula using cognitive linguistics (with definitions of the internet from magazines as corpus).

Seana Coulson gave a lecture on cognitive artifacts. She showed how a blending analysis can help in the analysis of cognitive artifacts (various lab equipment in an EEG lab).

Vito Evola presented PhD work on blending the erotic and the religious in mystical literature, with examples from The Song of Solomon.

Maurizio Gagliano presented his PhD work: Adapting Kant's schemas and Neisser's model of perception for a model of conceptual blending.

August 17

Topic: Methodology

Monica Gonzalez-Marquez played the latin song and participants attempted a blending analysis of the musical content and the lyrics.

Monica Gonzalez-Marquez presented experimental work on embodiment. Using portable and natural stimulus material (no computers), she aims at demonstrating, with high ecological validity, the psychological reality of image schemata.

Margaret Freeman gave a condensed talk about what readers do when understanding poetry. She illustrated inferences of varying complexity with examples from research on primates. No primates could spontaneously do a "systems mapping" inference, and according to her, neither could people she had tested either, with some execptions. But, she stated, people can learn how to do these complex mapping, necessary to understand poetry fully, so there is hope.

Special issues and questions discussed

How to go about practically when making a blending analysis

Often participants wondered about how to go about when making a blending analysis. How many input spaces? What to label them? On August 6, Rohrer gave some general hints for how to start making a blending analysis: 1) Brainstorm a list of lexical, visual, etc. items in the example to be analysed 2) Label some input spaces with labels that seem to capture the domains involved

What is a "mental space" in blending theory?

It was noted that some spaces are individual psychological phenomena while others are shared, social conceptualizations, not within any individual. An example of the former would be how a single person understands a particular text. An example of the latter would be how the concept of "complex number" has evolved historically.

How to determine what should be in any given mental space in the notational graph?

Some people proposed phenomenology as a basis for charting out the mental spaces (introspect to find what conceptualisation one has, and if there seems to be two distinct mental spaces or just one) - This view was brought forward by Line Brandt and Chris Sinha in the discussion. Seana Coulson disagreed and pointed out how often intuitions about cognitive processes are wrong.

Mixed questions to the faculty:

These are questions that were put in written form to the faculty in advance of the last day of summer school. These are their personal views on these questions.
  1. Would you say that blending theory makes any assumptions on the ontological status of consciousnesss? Is it consistent with all flavours of philosophy of mind, or deoes it require any special version?
    Replies:
    Margaret Freeman: It has the background of Lakoff & Johnson, therefore adheres to an embodied view of consciousness.
    Seana Coulson: Not per se, but people do.
    Tim Rohrer: Yes, assumptions on what consciosness is: embodied, dynamic. Some flavours are ruled out.

  2. Would you say that blending theory is compatible with both symbolic cognitive science (information processing) and connectionism? Why or why not?
    Replies:
    Seana Coulson: Computational models of blending have been symbolic. Both and neither; symbolic modelling is unpractical and connectionism, at the moment, provides too little structure.
    Tim Rohrer: The assumptions behind connectionism are the same as behind blending theory. There are hybrid systems, for instance "Neural theory of language".

  3. Would you say that blending theory deals with both conscious and unconscious cognitive processing (such as comprehending a common utterance - which would be automatic, unsconsciois - versus working out the solution to a complex problem, which require highly conscious reasoning)? Does it differentiate between the two?
    Replies:
    Seana Coulson: One of the things that people like the least about blending theory is that they are handled formally the same.
    Todd Oakley: Yes, it deals with both. Further work needs to be done.
    Tim Rohrer: Yes, it's a problem!

  4. It seems blending theory is used in two fundamentally different ways:
    1. To help the analyst understand the meaning of some piece of language for herself or himself (such as in literary criticism)
    2. Explain how a person in an authentic situation understands a piece of language (such as in empirical cognitive science)
    Do you think this difference is important to make? Why or why not?
    Replies:
    Margaret Freeman: I used blending theory first as a tool, but then ended up doing more than that... somehow it can model the conceptuals of the mind. Also, it is a great way for understanding another person's viewpoint.
    Todd Oakley: Blending theory is a good descriptive, pedagogical tool. It is an analytic, interpretative model: it started out that way.
    Tim Rohrer: Blending theory can be turned into (2), but it isn't right now.
    Seana Coulson: Even if (2) turns out to be wrong, (1) is still valuable. Blending theory hasn't been tested with psychological methods, but there is an article out soon on this topic by me [Seana Coulson].

  5. Challenge: Come up with an example of a cognitive phenomenon that cannot be handled by blending theory!
    Replies:
    Tim Rohrer: Blending theory should be used for creative examples only, not others.
    Seana Coulson: There are hugh numbers of examples! For instance, "why do people say 'uhm' in conversation?"
    Margaret Freeman: Becoming conscious of the pain of a mosquito bite.



Last modified Sept. 1, 2002 by Pierre Gander, pierregander@hotmail.com.