Processing models, a workshopA

David Bar-Tzur

Created 31 July 2001. Links updated monthly with the help of LinkAlarm.

processing areas in the brain

Materials needed by students:
One blank 120-minute videotape (VT), pens, and copies of the handouts listed below.

Materials needed by teacher:
handouts: Copies of this article itself, Colonomos activity, Gish activity, and Cokely activity.
videotapes: any appropriate texts to interpret during the activities.

Workshop summary

This workshop will discuss three processing models - Colonomos, Gish, and Cokely. By applying what you have learned about the Colonomos processing models, you will see what weaknesses you have in concentrating, representing, and planning. The Gish model will teach you how to listen for the structure of the text and thereby have a more organized input that your target audience can follow. The Cokely model will help you reduce the number of miscues (mistakes) you make, by seeing what miscue patterns you have, and why this is so.

What is processing?

A typist who takes dictation will listen to the message and parse it, that is, determine where the breaks in the sounds s/he hears are and which words are to be typed. This process usually occurs on a lexical level, that is, word for word. Sometimes the typist has to listen ahead to determine whether the phonemes "thai-er" is "there - in that place" "their - belonging to them" or "they're - a contraction of 'they are'", but for the most part each word is transcribed into its written/typed equivalent as it is heard. An interpreter, on the other hand, should wait for a complete unit of meaning, often a phrase, but ideally a sentence. This is because we are not transcribing from spoken English to written English. Even if we are using Sign English, we do not include every word, and must occasionally restructure the sentence we hear. The levels above "lexical" are called "phrasal" and "sentential". We shall give examples of what these levels look like. If an interpreter has a chance to prepare fully or is intimately familiar with the topic, s/he may rise to the textual level, being able to predict where the text is going and therefore be able to restructure the interpretation in a major way.

Depth of processing (English to ASL)

sL (source language, here spoken English): "Since there's nothing I can do, I might as well forget about it."

Lexical level of processing (word for word)


Phrasal level of processing (phrase for phrase)


Sentential level of processing (idea by idea)

tL: UP-TO-NOW ME STRUGGLE'over time'. SOLVE? (headshake). #DO-DO? PUSH-ASIDE-TO-left, TO-HECK-WITH-left.

Textual level of processing (taking into consideration the entire text and restructuring broadly)

The interpretation would restructure large portions of the text and the original sentence might be broken down and integrated into other sentences, so that you couldn't find one sentence that represents it.

Depth of processing (ASL to English)

sL: UP-TO-NOW ME STRUGGLE'over time'. SOLVE? (headshake). DO-DO? PUSH-ASIDE-TO-left, TO-HECK-WITH-left.

Lexical level of processing (word for word)

tL: "Up to now I struggle and struggle. Can I solve it? Uh-uh. Get rid of it. Darn it."

Phrasal level of processing (phrase for phrase)

tL: "I've been struggling with it. I doubt that I'll solve it. Just forget it."

Sentential level of processing (idea by idea)

tL: "Since there's nothing I can do, I might as well forget about it."

Betty Colonomos

Colonomos divides the interpreting process into three broad steps, which is very intuitive. (See "Pedagogical model of the interpreting process" on the next page.) The message must be taken in in the original language called the source language (sL), analyzed for meaning, and then produced in the language of the audience called the target language (tL). She calls the three processes that are needed to do this: concentrating, representing, and planning. In her older model of the interpreting process, the three steps were concentrating, visualizing, and rehearsing. Colonomos came to rename these two steps because the labels "visualizing" and "rehearsing" led to misunderstandings. The middle step does not have to be visual and we usually think of rehearsing as actually doing the task in order to improve it, rather than planning how to do the task. Notice that all three steps are very mental. The first is not called listening because the point is to describe what happens mentally (concentrating) in order to listen, everyone knows that the middle step must be mental, and the last step does not focus on the actual production of the message but the planning that comes before.

During concentrating, the interpreter seeks to understand the source message. While the interpreter listens, s/he must attend and analyze how what is being said relates to the event at hand. In representing, the interpreter must analyze the representation of the source message until it is stripped of language and remains as an idea or series of ideas beyond language. This representation can be a series of still pictures, an on-going movie, an abstract image, a feeling of movement within the body, remembered smells or sounds, and so on. In planning, this representation is now composed into the other language by assigning it words or signs. The interpreter must decide how to open and close sections of the communication, make transitions from one idea to the next, watch for red flags that may require more thought and requests for clarification, and finally deliver the target message.

What is involved in analysis of the source message and composition of the target message? First let's think of it without the presence of an interpreter. Any communication takes place within a context as seen by the black border around everything. Some aspects of the context are: (1) setting: is this in a church, psychiatrist's office, classroom, etc? Why is this important? (Do you chat in a church or praise the L-rd in a doctor's office?) (2) Participants: who are the people, what do they know about the topic to be discussed, and what are their relationships to one another, such as teacher-students or employer-employee? If you had to give a presentation on something you love, what would you like to know about your audience? (previous knowledge, attitude about the topic, are they all the same or not?) (3) Languages (ASL and English): How can I prepare myself for this assignment in terms of vocabulary or how can I become more masterful in ASL and English in general? (books, video tapes, dictionaries, encyclopedias, content knowledge, meeting speakers/teachers before an event/class, interacting with Deaf people, Toastmaster's) What influences a person's language? (age, gender, occupation, disability) (4) Culture: Is mainstream hearing culture and Deaf culture the only two cultures we might work between? (There is a greater emphasis on diversity, which means learning about other cultures such as Black, Hispanic, and Gay, which the Deaf person may be a member of.)

Colonomos divides the process skills needed for interpretation into analysis and composition. In analysis, the interpreter must derive meaning from the source message by attending and blocking distraction (side conversations, inner noise, heavy accent), storing information in short-term memory (names of people, dates, lists of things), accessing long-term memory for things already known by the interpreter to understand what was said conceptually. After the message has been broken down by analysis, the interpreter must use composition to restate the message in the target language using linguistic and cultural knowledge to state the message in a way that is comprehensible to the audience, access short-term memory for those items stored there during analysis, reaccess long-term memory to help plan the production of the target message, and then actually produce the interpretation. A speaker will often do this planning before s/he presents, and will tweak the message as s/he is monitoring him/herself and the audience, but sometimes s/he will actually pause and gather his/her thoughts.

Process management is what an interpreter does to accomplish analysis and composition when there are so many competing tasks to accomplish. The interpreter needs to pause at times to get a clear picture of what the message is (lag time), chunk related words or ideas together into a manageable whole, monitor if everything is going along smoothly, make decisions about asking for clarification or restatement of what the speaker (this includes signer) said, decide if a portion of the target message needs to be repaired, and decide if one's team interpreter needs to offer help.

It should be obvious that one must have as deep a knowledge as possible of both languages and cultures. This is an ongoing process and I am constantly learning about these for both languages and cultures. If I were to take a class in public speaking, for example, I would learn how to use English, even though I have an in-depth knowledge of grammar, syntax, vocabulary, and so on. Socializing with Deaf people teaches me more about their culture on a practical level, and now that diversity is playing a larger role in America, I need to begin to learn about Hispanic, Black, Gay, and Jewish cultures, to name but a few.

Content knowledge about the world (science, history, math, religion, and so on) helps the interpreter gain meaning from a speaker instead of having to both learn the material and "teach" the material simultaneously in the case of educational interpreting. Interpreters do pick this up as they do an ever increasing number of assignments: I recently did several automotive classes and now understand a great deal more about how a car works as well as the ASL signs that are used to express these ideas, but I can't always rely on waiting until the last moment. I need to prepare beforehand by making myself as aware and knowledgeable about the world as I can. This may be specific to an assignment (reading ahead in the textbook for a class one is interpreting) or general (learning new ideas before one "has to" interpret them).

Colonomos actually takes this idea of preparation and expands it a good deal by talking about such things as getting enough sleep and staying healthy, cultivating confidence, managing stress, meeting with the speaker beforehand, learning who all the participants are and what the environment will be like (room, outdoor area, auditorium), and meeting with team members and coordinators. There is also an internal environment inside the interpreter that can lead to "noise", that is thoughts or feelings that nag at the interpreter while s/he is trying to work: hunger, fatigue, illness, repulsion at the topic, lack of confidence. Working on minimizing these will lead to a better interpretation.

Do the Colonomos activity.

(1) Please write your name on your videotapes if you have not already done so. (2) Break up into groups of three. (3) One person will tell a 3-5 minute story. It can be about anything, but here are some ideas: "An embarassing moment", "An aggravating person", "A pleasant time I spent with a friend", "How I first met my 'significant other'", "How I became interested in my favorite hobby/sport/pasttime". (4) The second person will work the camera and watch the time. (5) The third person will give feedback on what is seen by using the Colonomos activity form. (Pass them out.) Remember to write down the name of the interpreter you are observing on the form so that they can read it again later. (6) When the story and interpretation is done, stop the tape and the interpreter will talk briefly about how they felt about their interpretation. (7) The third person will use the feedback form to remind them of what they saw on the various categories. The categories are: (A) How did the interpreter do at concentrating? Large portions omitted or sentences not finished? (B) How did the interpreter do at representing? Did s/he provide a clear picture of visual things and use space to show the relationship of things and ideas? (C) How did the interpreter do at planning? Did the message look well-planned or disconnected? (8) When all the interpreters have finished receiving feedback everyone should meet again as one group.

Ways to improve in the future if you are weak at -

Concentrating: Ask yourself what distracted you: (1) trying to work too hard on one part of the message while the other parts were still coming at you, (2) external things in the environment (movement, extraneous sounds, the speaker's style, too hot, etc), (3) internal "noise" (not enough sleep, hungry, outside problems in the real world, inner doubts about skills, etc), (4) lack of skill at concentrating on any task (an on-going problem for things other than interpreting also). Come up with a plan that will minimize this sort of thing in the future. If (1) above, try to let go of things you can't do at present or combine ideas from one part of the message with other later parts. If (2) ask for the environment to be changed if possible, such as asking for participants to not talk while the speaker is talking, have the lighting changed, move, bring a sweater, etc. If (3), make sure you do get enough sleep, eat well before an assignment, work on your confidence through self-help books. If (4), seek medical or psychological help.

Representing: Experiment with different kinds of representing. If you usually use still pictures, use movies; if visualizing is not actively done, try it now; if you never thought of using remembered smells or sounds, do it. Make the message become so real to you that you sense it happening in front of you. Allow yourself more lag time so that you can get a complete picture of what is happening.

Planning: Listen to a tape-recorded message and try to plan it out on paper. By this I mean, make an outline of how you will put things in space (if it is English to ASL), any contextualizing (explaining the concept more fully than may be done in the original, such as "mainstreaming" is not understood by most hearing people), any reorganization of the text, referring back to previous sections, repetition that is not in the original, and so on. If you are going from ASL to English, you may need to plan how you will put classifiers into English words. Stop the tape at first when you need to, and then try doing it with the tape-recorder running. Listen to the tape-recording again and see if you can improve on it. Allow yourself more lag time so that you are planning for more than the next word.

Sandra Gish

The term "model" is actually used for two different things: the service models (machine, conduit, communication facilitator, ally) and the processing models. Gish's is not exactly a service model, although you could say it is in so far as it has us become "information processors." Cokely definitely has his complete flowchart of what happens in an interpreter's mind although you could argue that it still doesn't say "how" this happens. Although his work is very scholarly, and makes a good case for classifying miscues as he does, I don't see how that verifies that his flow chart is in fact how our mind processes a message for interpretation.

Colonomos has a flowchart for all the influences that impact us as interpreters and seems more useful as a pedagogical model and seems intuitively valid. With Gish I would say that she does not so much tell us how the mind processes information as how we should structure it so that the mind can process the information in a way that will lead to a semantically equivalent interpretation. Since this deals with how we can better process the message, it seems reasonable to call it a processing model. Gish is a beautifully humble soul and would not be crushed if someone called it an approach rather than a model.

As we have said, Gish's "model" uses an information processing approach. We can use computer translation as a paradigm. There are computer programs that will translate from one language to another. The problem with them at present is that they do one word at a time, and any given word may have many meanings. The most famous example is the word "run". Does it refer to moving your legs quickly while you move from place to place, an ever-increasing tear in nylon stockings, competing with other people for political office, an ongoing stream of mucus from one's nose, the sudden onrush of people to do the same activity (as in "run on the market")? Computer programs will presumably be designed in the future that will take into consideration the whole sentence or better yet the whole text. This need to understand the whole text is what drives Gish to the representation you see in the following overhead.

Rows showing (1) goal, (2) theme, (3) objectives, (4) sub-objectives, (5) units, and (6) details.

A lengthy text will have a goal (to convince, to educate, to motivate), a theme ("The history of public education"), objectives (to show the progression of ideas), sub-objectives, units, and details which all are finer and finer division of the ideas that when summed up will lead to an understanding of the history of public education (for example). In the last two rows of the diagram above, you see a series of boxes connected to one another by lines. The upper row has boxes labeled "unit" and the lower row "detail". The units could be thought of as a paragraph or series of paragraphs, and the details as sentences or phrases. These two rows of boxes are further connected from above to dashed-line boxes by lines all the way up to the "Goal". The dashed lines show that the speaker may not explicitly state what the goal is. It is helpful when interpreting to ask yourself, what is it that the participants have gathered together to do? Especially, what is the goal of the speaker?

Sometimes the goals of the speaker and the audience are not the same, as when a student has been called to the principal's office. The principal's goal is to punish behavior and the student's goal is to escape punishment. The dashed lines around "Goal" and the three rows beneath it show that the speaker may not explicitly state these things. The principal will not announce his goal saying, "I have called you into my office to punish you." This is well known to the student. Nor will the student say, "What I will now explain is meant to avoid punishment." This would obviously be foolish. Other communication goals are: to convince, to comfort, to confuse, to seek revenge, to teach, to learn, to declare one's love. The list is endless, but the point is to determine what is being attempted here. Usually you who are K-12 interpreters are working with the goal of "to teach" although other goals come up, such as in the principal's office.

Now you will need to determine what exactly is the teacher trying to educate the students about. This is the Theme, which is the second box down on the overhead. For example, the teacher wishes to demonstrate that Black people have been oppressed. This will lead to a series of objectives as shown in the next row down. In the introduction the teacher might speak of what oppression means in general and how Africans were brought to America through the slave trade. In the body of her presentation, the teacher may speak of the treatment of slaves before the civil war as one objective, how they were treated after the Civil War until the Civil Rights movement as another objective, and how oppression is still exemplified today as the final objective. In the conclusion, she may restate the theme and summarize what you can learn from what has been stated. These objectives may have sub-objectives, but we shall not deal with these. The more you can know as an interpreter about what will be said and how, the better you can interpret in a way that will have an equivalent impact on the target audience. This can be done by preparation, content knowledge and analyzing what is being said as it happens.

We have spoken of the objectives, now we will deal with the units. It could be that the teacher divides the way that slaves were treated before the war into the following units: (1) prohibition of the use of the slave's native language and the imposition of English as the only permissible language, (2) prohibition of the use of the slave's native religion or spirituality and the imposition of a specific denomination of Christianity (probably evangelical) as the only permissible religion, (3) the destruction of the nuclear family as the master decides who marries whom and split families up as it suits the slave-owning family, and (4) the forced labor of slaves without due remuneration. These units will have details within them that are represented by the bottom row of the chart.

The interpreter should chunk together all the details that fit under one unit and show the change from one unit to the other. The listing of details can be shown by indexing the fingers on the hand (point to index finger, then the ring finger, and so on). The change from one unit to another can be shown by signing (2h)NEXT-ON-LIST (both bent B hands with palms facing moving downwards). As mentioned before, the top four rows on the chart may not be stated explicitly, but the audience members who share the language used by the interpreter should come away with the same understood goal, theme, and objectives as the audience members that share the same language as that used by the speaker/signer.

Do the Gish activities. Identifying speaker goal and theme

(1) Get into three groups. Pass out Gish activity 1. It can be the same groups or not, your preference. Pick a videotape in ASL to show the students. Ideally it should be a long text so that it can be used for each of the parts of this activity. Before viewing the tape, have the students analyze the title or topic. I use Forestal, E. Understanding the dynamics of Deaf consumer-interpreter relations. Front Range Community College Interpreter Preparation Program. If you feel that the original title of the tape you use isn't very revealing, give it one of your own. Students: Think about the reason this speaker is talking. What do you think her goal is?
(2) I will start the tape and then freeze-frame the speaker. How does the gender, age and overall appearance of the speaker influence the prediction of the speaker's goal?
(3) I will show you a few minutes of the tape. Watch it for understanding. Do not attempt to "interpret" it in your head. Don't be intimidated if you don't understand everything: we are looking for the main point. When you have finished watching it, you will write down one or two English sentences that you feel summarize the entire portion you watched. Everyone write this down without discussion.
(4) Now each of you takes turns signing your sentence(s), no more than two or three that summarize what was said.

Gish activity 2: Identifying units and details

(1) (Hand out paper for students to write on.) I will play a five-minute segment of this same videotape. For now I will have you listen to the English and not watch the ASL. While the tape is playing, everyone in the group should take notes. The goal is to come up with a short outline of what was said. One of you will use this outline while you are interpreting a replay of this segment.
(2) I will now replay this segment and one of you should interpret this segment using your outline as a guide. This time you will be watching the ASL and the English interpretation will be turned off. Because you have an outline you should not have to struggle so much with the ASL. Try to show in your interpretation, when a new unit is started and include all the details in that unit. The others in the group should take notes to give feedback to the interpreter after the segment is through.
(3) When the segment is done, the interpreter can express how s/he felt his/her work (interpretation) went. The others may then give their feedback. Focus only on if the units and details were clearly expressed, separated and not run together, and whether any were skewed or missing.
(4) This process is repeated until everyone has had a turn.

Dennis Cokely

This model is presented at length in Interpretation: A sociolinguistic model (1992). In this study Cokely sets out to go beyond the information processing models that had been proposed by Ingram (1974), Gerver (1976), Moser (1978), and Ford (1981). He proposed a model that would include not only mediating between two individuals and communities (cultures?). It seems that Cokely's model was constructed by performing an extensive analysis of 6 interpreters who were working from English to ASL during a Conference of Interpreter Trainers convention. Their miscues were categorized and Cokely used these to construct a sociolinguistic model.

A miscue is a "lack of equivalence between the sL (source language) message and its interpretation or, more specifically, between the information in an interpretation and the information in the sL message it is supposed to convey." (p 74) There are five possible miscues: omissions, additions, substitutions, intrusions and anomalies.


sL: "What do I mean by these policy decisions?"

back translation: "What do I mean by 'policy'?"
omission: The focus is on the decisions, not the policy.
better tL: POLICY DECIDE"each" "WHAT"?

Omissions may be: (1) morphological - where content information that is clearly contained by bound morphemes such as "teachers" being signed TEACHER without a THEY or GROUP to show plurality, (2) lexical - involving a word or short phrase, or (3) cohesive - where the relationships of ideas is neglected. For example, if the English were: "He's shy, because he doesn't know you" and was interpreted HE SHY. HE NOT KNOW YOU. Really the sentence should be interpreted NOW+ HE SHY. WHY? GO-AHEAD KNOW YOU, NOT-YET.


sL: ". . . an analogy to the simultaneity of listening and speaking in simultaneous (interpretation) . . ."

tL: 	             __________________________________th

back translation: ". . . an analogy to carelessly listening and speaking at the same time."
addition: carelessly (th).


Additions may be: (1) nonmanual - where these are added because the interpreter thinks that this is part of the sign, such as th above, (2) lexical - where a word or phrase is added to contextualize (explain the context) but may be misleading or unintended, or (3) cohesive - relating ideas in the text that are not meant to be related.


sL: "If I was studying French history. . ."

back translation: "If I were studying French culture"
substitution: "culture" rather than "history".

Substitutions may be: (1) expansive - including more than was intended as above, (2) restrictive - including less than is signed for "more importantly I have to decide", or (4) unrelated - ENGINEER is signed for "plumber".

(Source language) intrusion A source language intrusion is when the source language (the language that the original message was produced in, be it English, American Sign language or any other spoken or signed language) influences the interpretation, usually in an unnatural choice of words or a grammatical error that makes sense in the source language, but not in the target language.

sL: "If you get my meaning. . ."

back translation: "If you obtain my meaning." intrusion: the English word "get" can mean "understand", but the sign GET can not.

An intrusion may be: (1) lexical - where the gloss for a sign is overgeneralized - as in GET above or (2) syntactic - where the interpretation is structurally too English and the ASL user has to reconstruct what the original English must have been and reinterpret it him/herself. The example above is also syntactic and needs to be restructured as shown above.


An anomaly "refers to instances in which the tL message is meaningless or confused and cannot be reasonably accounted for or explained by another miscue type." (p88)

sL: "The matiere courses were taken in other departments."
tL: SECOND-THING OTHER NEXT INDEX-'first and second item on list of five' M-A-T-U-R-E THAT INDEX--'second item on list of five' OTHER P-T-S.
back translation: "Second item of the next list of items first item second item mature that second item other p-t-s."
anomaly: no verb and reference to a list that was not set up previously.

Omissions tend to be the most frequent in occurrence. This may be due to processing interfering with concentrating on everything that is heard. Not all omissions are equally bad. A lexical omission of some items from a list where enough items are conveyed to give the general idea is not as bad as a cohesive one that omits the name of the topic that will be discussed next. The consumer may be able to recover the information because the speaker is redundant in his/her presentation, or the consumer knows the information at hand is wrong and can substitute knowledge of the truth for what was said/signed.

Cokely's study showed that interpreters with longer lag time, even by a few seconds, had far fewer additions, presumably because they understood the message and did not need to try to anticipate (often wrongly) what would be said next. Just waiting longer is not the solution to everything of course, the interpreter must be able to retain what was heard. Too short of a lag time will lead to an over literal (too English) interpretation which must be fixed by adding a correction of what was missigned (or misspoken). Often there is no time for a correction, the interpreter may be embarrassed to show his/her error or may feel that a correction would be intrusive. One of the interesting additions that was found was that of nonmanuals. I often observe interpreters add th (the tongue stuck out) or mm (the lips pursed like a kiss) in inappropriate places as was found in this study. It could be that interpreters think certain signs always get this nonmanual or they may be trying to look Deaf by not mouthing every word and use these nonmanuals as something to do with their mouth.

Substitutions, as with additions, can not always be recognized by a consumer, and the message can not be easily recovered. Restrictive substitutions are not as bad as expansive ones, because at least they do not include too much. Cohesive substitutions are often worse than both, because they may lead to a misunderstanding of temporal order, causality, or whether an extended idea is to be included or excluded.

The degree to which intrusions are problematic will depend on the extent to which the Deaf consumer can understand Signed English. The consumer him/self may sign in non-conceptual ways because of learning ASL from hearing people or trying to appear educated. The interpreters in the study that had a shorter lag time had three times the intrusion error of those with a longer lag. Intrusions may erode the trust that a consumer has in the interpreter. This will create psychological noise that interferes with the consumer's concentration on the topic at hand and so will the reinterpretation that happens as the Deaf person tries to reconstruct what the English must have been and reinterpret it.

Anomalies are the easiest for a Deaf consumer to notice, since they result in a message that simply doesn't make sense. Here the message will be the least recoverable. Lag time does not seem to have a clear relationship to anomalies. Another miscue that was briefly dealt with was that of style or register. The Deaf delegates to the convention felt that half of the interpreters were much less formal than the other half and Cokely found this reflected in the observations of other Deaf people to whom he showed the videotapes. This would lead to audience members discounting the authority of a given speaker, such as you might feel if you went to a professional convention and a speaker kept using the word "ain't". In general, those interpreters who used more nonmanual grammar (topic, rhetorical questions, wh-q and q) were deemed more formal. Also there was less misuse of "th".

To be fair we have to recognize that speakers also have miscues. If you have ever read a transcript of a spontaneous speaker, you will be amazed at the number of grammatical errors and rephrasing that take place: "Okay what I want you to do, I. . . we're not. . . there's gonna be certain information." We have learned to filter these out and simply listen for the message but if you really stop and analyze what was said the errors become glaring. The same thing happens when you watch a taped interpretation.

Cokely found that 16% of all interpretations were anomalous, that is meaningless or confused. Some of the reasons for this are: nervousness at being in so formal a setting, not being able to control the pace of the speaker, inability to prepare as well as might be helpful, an attempt to suppress PSE, a lack of exposure to formal ASL. Distributing the other miscues over the sentences that did make sense, it was found that there were 1.21 per sentence. That is, almost every sentence had at least one mistake if not more. And this is with highly skilled conference interpreters. When the more serious miscues were considered, that is the ones that were difficult to recover, even knowing English and the content material, .57 miscues were found per sentence. Much of the message is not recoverable. We must bare in mind that the message was very complex, and the interpreters probably did not understand all of it themselves. This is not meant to make you feel that there is no point to interpreting anything, but that if you don't increase your skills, almost nothing will be left to be understood by your client.

Cokely activity:

(1) Form groups of three people: 1 speaker, 1 interpreter, and 1 observer/camera person.
(2) The speaker will present for 5 minutes on some topic with which s/he feels comfortable. The interpreter will interpret this into a video camera using his/her videotape labeled with the interpreter's name. The observer will start the camera, make notes on the interpretation, and stop the camera.
(3) The interpreter for that session gives feedback on her/himself first and then the observer uses his/her feedback form to relate to the interpreter what miscues s/he observed.

Let's look at the Cokely processing model step by step and talk about how miscues happen and therefore might be minimized. The box furthest to the left shows auditory and visual input being feed into a box to the right and one above. The one above is called "message reception". We will clearly want to have the best environment that we can: good lighting without glare, good sightline and distance, good acoustics and distance. Good glasses and hearing aid if necessary (I'm starting to use one). Struggling to hear or see will of course lead to miscues as we try our best to guess what we have heard or seen. This will also require you to sometimes ask for repetition of what was said.

In the box below "message reception" we see the "sensory register": our eyes, ears and what connects them to the brain. Next is "recognition", which requires us to understand the words/signs we are taking in as well as other aspects of the language. If we don't understand the content of what is being said, we will struggle with the meaning. Just because you know English fluently doesn't mean you would understand a lecture on quantum mechanics. Below this is "attention", which requires us to develop the skill to attend to what is being said, ignoring internal reactions to the message, to the people around us, and to our own problems, whether they are because of the job at hand or troubles in the outside world.

The next column to the right is "preliminary processing" where the words heard and the signs seen are understood. In fact it requires context to really understand the words/signs. We bear in mind: what is the goal, where are we, and what do we know about the participants? Think how hard it is to hear a list or unrelated terms and really listen for very long. Below this box is "influencing factors". Our "lexical bank" is all the words we know with all their shades of meaning. "Syntactic competence" helps us both understand words through their position in a sentence as well as predict what might be coming next. "Source language long term store" is the sum total of everything we have learned over the years about the source language being used.

The next column to the right is "short term message retention". We will hear details that we need to remember long enough to convey them. The more we already know about what is being said, the less we have to remember. The more we prepare for an assignment the more we can retain because we already "know" it. We also have to remember a phrase and possibly a sentence before we can make sense of the single words, some of which may have only partially been heard or attended to. If a person says "ra-it", was that "riot", "right (not left)", "right (not wrong)", "right (to do it)", "write", or "rite". We need to hear more and remember that series of sounds or remember what was said before.

These units we hear, remember, and interpret are called chunks, and the process is called chunking. These chunks can be remember and built into larger chunks. In order to chunk we need to wait for understanding, called lag time. Miller (1956) observed that our short-term memory is capable of retaining 7 + or - 2 items. Ideally these items should not be words or most sentences could not be memorized, better if the items are concepts that are connected together. Below this are the "elaborative functions" that allow the ideas expressed by the words to be combined into concepts or ideas and remembered as such. "Rehearsal" is what happens in the mind as we try to compose the sentences that we are about to utter. We can see that full sentences are composed before we start them by "spoonerisms" we sometimes say, such as "it's kiss-tomary to cuss the bride" (where "kiss" and "cuss" are inverted) instead of "It's customary to kiss the bride". And "decay" is what happens as items in short-term memory are forgotten.

Next to the right and above is "semantic intent realized", where a portion of the message is now understood and the process of formulating it into another language can now begin. Cokely gives us a wonderful example of how this can go wrong. The sL message is "We [sign language interpreters and spoken language interpreters] testify with one voice and not jeopardize our standards [of interpretation that we wish the law to specify]." The interpretation is TWO-OF-US lf-FIT-IN-rt ONE VOICE AND NOT MAYBE HAPPEN DESTROY TWO-OF-US WORK TOGETHER. This can be back translated as follows: "We can merge into one voice and not jeopardize our working relationship." This misinterpretation came about because the interpreter did not understand "jeopardize standards" which is a bit unclear. Because the interpreter heard "jeopardize" s/he anticipated that the speaker meant to say "working relationship" next and didn't allow for lag time to process the meaning.

The next three top boxes show what happens once the sL message is understood. In "semantic equivalent determined" the mind is considered how to express the message in the tL. In "syntactic message formulation", the mind makes sure that the interpretation is grammatical, and then in "message production" the interpretation is expressed in either oral or signed output. Below these three boxes are the factors that operate during these stages, including linguistic and cultural competence. Below all of these, each competence is broken down into sub-skills. Just as we think about how we want to express our own ideas in the best way possible, we need to listen, understand, and then rehearse in our minds the best way to express that idea.

This workshop has discussed three processing models - Cokely, Colonomos, and Gish. We will also talk about the concept of a miscue. For now we can just say that a miscue is a mistake that is consistently made due to some breakdown in the processing of the message. By applying what we learn about the different processing models, the students will see what sort of miscues they make and become increasingly able to diminish the number of miscues produced over the span of their career by identifying their strengths and weaknesses and producing cultural linguistic equivalents. Today we will use videotapes to practice interpreting from English to ASL and ASL to English towards that end.

Workshop summary

This workshop has discussed three processing models - Colonomos, Gish, and Cokely. By applying what you have learned about the Colonomos processing model, you will see what weaknesses you have in concentrating, representing, and planning. The Gish model will teach you how to listen for the structure of the text and thereby have a more organized input that your target audience can follow. The Cokely model will help you reduce the number of miscues (mistakes) you make, by seeing what miscue patterns you have, and why this is so.


Cokely, D. (June 1992). A sociolinguistic model of the interpreting process: ASL & English. Linstock Press. ISBN: 0932130100.
Colonomos, B. M. (1989). The interpreting process: A working model. Manuscript.
Gile, D. (1995). Basic concepts and models for interpreter and translator training. Philadelphia, PA: John Benjamins Publishing Company. ISBN: 1556197039.
Gish, S. (1996). The interpreting process: Introduction and skills practice, the Gish approach to information processing.
Lee, R. G. Cognitive processing: The brains behind the work. Signs of Development, Inc. "This workshop provides an overview of the interpreting process as well as a detailed description of cognitive models of interpreting. Areas covered include: Necessary skills for interpreters, communication as a process, general overview of models, descriptions of the Colonomos and Cokely models, the role of monitors in the interpreting process. To order.

Image credits

A. Center for Biomedical Imaging Computing - Johns Hopkins University School of Medicine, Department of Radiology.