We know that for second language acquisition to occur students need to hear or read meaningful input. If the message what they hear or read is not understandable you might as well expose them to gibberish. In fact, there has been research into how many new words students can cope with while maintaining a meaningful message. Don't forget that students can use compensatory strategies, e.g. their knowledge of the world, their hypotheses about what a text might mean, their knowledge of cognates and so on, to work out meaning to some extent.
Studies, for example those carried out by Paul Nation, indicate that for a text to be understood a bare minimum 90% of the words need to be already known. For most learners this figure rises to around 98%, maybe even more. If you include more unknown words than that students lose the message (and may switch off as a consequence).
Those percentages may seem high to you, so to test them for myself I have taken a text from frenchteacher.net and looked at the percentage of vocabulary which I glossed, based on my assumptions of how much a typical A-level student would know. (This is problematic, as I'll refer to below.)
So, without copying the entire text here, I took a first text I just uploaded about homelessness in Quebec. The text word count is 370 words, as calculated by MS Office. My word/short chunk glossary totalled 14 words. In theory this text is 96% comprehensible by this crude definition and what I assumed the typical student to already know.
I then took a second text, a fait divers of intermediate level (Higher Tier GCSE) about an archeological discovery. This text contains 300 words and I glossed 27. In this case therefore the "understandable" count is 91%. I actually compensated for this, as it happens, by including an English-French 'find the French task, realising that the text was just a tad too hard.
Now, these figures are a reflection of my own assumptions about how much help I needed to give to make a text meaningful, interesting and teachable. It seems like my own gut feeling is broadly in line with what research says. But there are some serious caveats and observations here! Here are a few I can think of:
1. Role of syntax and morphology
Understanding a text requires more than just knowledge of words and lexical chunks. Students have to be able to 'parse' sentences to make meaning too. If their grammar is poor they will make mistakes. This often means, when writing texts, that you need to simplify the grammar as well as the vocabulary, prioritising higher frequency and usefully transferable words, chunks and structures. To make a text more comprehensible in this regard a glossary can include conjugated verbs or longer chunks rather than infinitives and isolated words.
2. Individual variations
Each student knows a different number of words so when you write or choose a text you have to arrive at a Goldilocks-type compromise: not too hard, not too easy. On the whole you might want to lean towards the easy end to help all students as far as possible. Comprehension is the key. So you cannot guarantee 90% + comprehensibility for all students.
3. Cognates
They help a lot with languages like French, Spanish and German, but students' mastery of English varies so we cannot always rely on them.We sometimes assume a student knows the word in English when they don't. In addition we have the occasional issue of 'false friends' which can create misunderstandings.
4. Compensatory strategies
The background knowledge of the world and discourse construction affects how much a student understands. Two students with equal mastery of vocabulary may perform differently based on their general or specific background knowledge. Some will also be better at working out meaning through their grasp of discourse structure, e.g. how arguments are constructed.
5. Parallel texts
These can compensate for a lack of vocabulary or grammatical skill by providing instant translation and access to harder content. However, is this the equivalent of watching a film with L1 sub-titles? Is this the most desirable form of comprehensible input? I would say not, since students are not applying their existing knowledge of vocabulary and syntax to any great degree. they are not building a little on what they know, a generally sound principle of learning.
6. Exercise type
One way of compensating for lack of vocabulary is to provide questions in English for comprehension. This fulfills a similar function to parallel texts in that you are giving new vocabulary through the questions. Again, you may feel this is less than ideal, partly because you have to resort to using English when you might prefer to be working on the target language as much as possible.
7. Self-efficacy
The feeling of self-confidence, belief in one's ability to do well, is vital for maintaining motivation and progress. When you expose students to texts they just barely understand, they are likely to feel they are failing. We want students to feel successful - many believe this is the number one factor in motivation.
8. Overestimating what students know
Although I taught for 33 years and should have learned from experience, I think I probably always assumed that students knew more than they did. We language teachers are the successful linguists who may have found it all relatively easy compared with the bulk of students in our classes, many of whom find it really hard! Maybe we sometimes pitch texts too high or don't provide enough scaffolding to make them comprehensible. Just a thought.
These points raise interesting questions for language teachers. For example, with regard to the role of authentic texts. Some teachers vociferously support the use of authentic texts, arguing that they are more motivational and provide better reflection of the TL culture. My observation here is that an authentic text is fine when it is at the right level, but that comprehensibility is more important than authenticity. In addition, concocting texts or adapting authentic ones allows the teacher to tailor the input more precisely to the class - taking advantage of what they already know (recycling) and allowing you to provide built in repetition within a text. This also allows you to personalise the content to the class, to the point of including references to the pupils, school or teacher.
Secondly, written texts are easier to cope with than spoken texts owing to the added issues of having to identify words in the sound stream. Frequently a student may know a word but simply doesn't recognise it between others in the flow of sound. Different accents and quality of sound reproduction may not help either. So this is an argument for making listening texts even more comprehensible than written texts, or at the very least finding ways of making them easier to understand.
Lastly, I would not want to come across as too dogmatic about how meaningful texts must be, whatever the research suggests. There is surely a case for letting students occasionally loose on tougher texts with the help of a dictionary or 'adapting the task to the text', as they say. If you are teaching a novel or film for A-level you are forced to engage students with language considerably above their level. In my view, however, this is generally a second-best solution.
If you would like one research reference on this issue of comprehensibility try this from Paul Nation (go to page 61):
https://www.victoria.ac.nz/lals/about/staff/publications/paul-nation/2006-How-large-a-vocab.pdf
Studies, for example those carried out by Paul Nation, indicate that for a text to be understood a bare minimum 90% of the words need to be already known. For most learners this figure rises to around 98%, maybe even more. If you include more unknown words than that students lose the message (and may switch off as a consequence).
Those percentages may seem high to you, so to test them for myself I have taken a text from frenchteacher.net and looked at the percentage of vocabulary which I glossed, based on my assumptions of how much a typical A-level student would know. (This is problematic, as I'll refer to below.)
So, without copying the entire text here, I took a first text I just uploaded about homelessness in Quebec. The text word count is 370 words, as calculated by MS Office. My word/short chunk glossary totalled 14 words. In theory this text is 96% comprehensible by this crude definition and what I assumed the typical student to already know.
I then took a second text, a fait divers of intermediate level (Higher Tier GCSE) about an archeological discovery. This text contains 300 words and I glossed 27. In this case therefore the "understandable" count is 91%. I actually compensated for this, as it happens, by including an English-French 'find the French task, realising that the text was just a tad too hard.
Now, these figures are a reflection of my own assumptions about how much help I needed to give to make a text meaningful, interesting and teachable. It seems like my own gut feeling is broadly in line with what research says. But there are some serious caveats and observations here! Here are a few I can think of:
1. Role of syntax and morphology
Understanding a text requires more than just knowledge of words and lexical chunks. Students have to be able to 'parse' sentences to make meaning too. If their grammar is poor they will make mistakes. This often means, when writing texts, that you need to simplify the grammar as well as the vocabulary, prioritising higher frequency and usefully transferable words, chunks and structures. To make a text more comprehensible in this regard a glossary can include conjugated verbs or longer chunks rather than infinitives and isolated words.
2. Individual variations
Each student knows a different number of words so when you write or choose a text you have to arrive at a Goldilocks-type compromise: not too hard, not too easy. On the whole you might want to lean towards the easy end to help all students as far as possible. Comprehension is the key. So you cannot guarantee 90% + comprehensibility for all students.
3. Cognates
They help a lot with languages like French, Spanish and German, but students' mastery of English varies so we cannot always rely on them.We sometimes assume a student knows the word in English when they don't. In addition we have the occasional issue of 'false friends' which can create misunderstandings.
4. Compensatory strategies
The background knowledge of the world and discourse construction affects how much a student understands. Two students with equal mastery of vocabulary may perform differently based on their general or specific background knowledge. Some will also be better at working out meaning through their grasp of discourse structure, e.g. how arguments are constructed.
5. Parallel texts
These can compensate for a lack of vocabulary or grammatical skill by providing instant translation and access to harder content. However, is this the equivalent of watching a film with L1 sub-titles? Is this the most desirable form of comprehensible input? I would say not, since students are not applying their existing knowledge of vocabulary and syntax to any great degree. they are not building a little on what they know, a generally sound principle of learning.
6. Exercise type
One way of compensating for lack of vocabulary is to provide questions in English for comprehension. This fulfills a similar function to parallel texts in that you are giving new vocabulary through the questions. Again, you may feel this is less than ideal, partly because you have to resort to using English when you might prefer to be working on the target language as much as possible.
7. Self-efficacy
The feeling of self-confidence, belief in one's ability to do well, is vital for maintaining motivation and progress. When you expose students to texts they just barely understand, they are likely to feel they are failing. We want students to feel successful - many believe this is the number one factor in motivation.
8. Overestimating what students know
Although I taught for 33 years and should have learned from experience, I think I probably always assumed that students knew more than they did. We language teachers are the successful linguists who may have found it all relatively easy compared with the bulk of students in our classes, many of whom find it really hard! Maybe we sometimes pitch texts too high or don't provide enough scaffolding to make them comprehensible. Just a thought.
These points raise interesting questions for language teachers. For example, with regard to the role of authentic texts. Some teachers vociferously support the use of authentic texts, arguing that they are more motivational and provide better reflection of the TL culture. My observation here is that an authentic text is fine when it is at the right level, but that comprehensibility is more important than authenticity. In addition, concocting texts or adapting authentic ones allows the teacher to tailor the input more precisely to the class - taking advantage of what they already know (recycling) and allowing you to provide built in repetition within a text. This also allows you to personalise the content to the class, to the point of including references to the pupils, school or teacher.
Secondly, written texts are easier to cope with than spoken texts owing to the added issues of having to identify words in the sound stream. Frequently a student may know a word but simply doesn't recognise it between others in the flow of sound. Different accents and quality of sound reproduction may not help either. So this is an argument for making listening texts even more comprehensible than written texts, or at the very least finding ways of making them easier to understand.
Lastly, I would not want to come across as too dogmatic about how meaningful texts must be, whatever the research suggests. There is surely a case for letting students occasionally loose on tougher texts with the help of a dictionary or 'adapting the task to the text', as they say. If you are teaching a novel or film for A-level you are forced to engage students with language considerably above their level. In my view, however, this is generally a second-best solution.
If you would like one research reference on this issue of comprehensibility try this from Paul Nation (go to page 61):
https://www.victoria.ac.nz/lals/about/staff/publications/paul-nation/2006-How-large-a-vocab.pdf
Comments
Post a Comment