Jennifer A. Lacroix, Boston University, Boston, Massachusetts, USA and Abigail J. Castle, University of Iowa, Iowa City, Iowa, USA

Bottom-Up Versus Top-Down Listening Processes

In cognitive linguistics, the process of listening comprehension is often referred to as top-down vs. bottom-up listening. A way to conceptualize bottom-up processing is through the lens of an L2 learner who must attend to intonation patterns, word prefixes, or other linguistic features to make meaning more accessible. An example of bottom-up processing is understanding each word and detail (i.e., attempting to figure out the meanings of most words or sentences in the input)—what the literacy world calls decoding. When L2 listeners are decoding acoustic information, the decoding process starts from the sound elements of the target language, such as phonemes and syllables, and then progresses into words, phrases, and sentences. In tandem, listeners also use top-down processes which tap into prior knowledge and enable listeners to use strategies such as note-taking and inferencing to make sense of speech.

A more advanced knowledge on this topic includes the notion that neither of these processes happens in isolation and that they are rather an interactive process (Field, 2008). The challenge, in our view, as seasoned practitioners who have taught EAP classes for many years and conducted our own research, is that often the bottom-up skills are overlooked in EAP contexts in favor of a more traditional academic approach, which favors note-taking as the common academic listening practice. What note-taking presupposes, however, is that the L2 learner already knows how to make sense of the stream of speech. That is, he or she can already recognize presumably known words in their aural form and do something with that knowledge (Siegel, 2016).

A number of studies have investigated the effects of bottom-up listening processes. For example, Goh (2000) has identified problem areas in listening perception and parsing, concluding that problems with perception were greater than parsing. Matthews and Cheng’s (2015) experiments with word recognition also relate to one of the perception problems that Goh (2000) identified, which was the inability to recognize known words in connected speech. Jensen and Vinther’s (2003) repetition studies further tried to find a solution to these perception and parsing problems and found favorable evidence for the role of exact repetition. All of these studies give insight into what an evidence-based practice that integrates theory into practice looks like.

Our view is that by reconceptualizing our L2 listening pedagogy to—once again—include these bottom-up processes, L2 listeners can more consciously and holistically work toward advancement in their L2 listening comprehension skills.

Five Explicit Strategies Identified in the Research

The following explicit strategies are intended to raise learners’ metacognitive awareness of the speech stream and provide opportunities for prediction and reflection. Additionally, these strategies allow learners to rely on their other English language skills to bolster their listening proficiency.

1. Dictation

Dictation is a useful practice that allows learners to rely on learned sentence structures and vocabulary. This activity should emphasize content and raise awareness of function. In this activity, sentences may be dictated multiple times and should be stated with standard word stress, sentence stress, intonation, and thought group pauses. The length and complexity of the sentences is determined by proficiency level and curricular goals. Grading criteria and/or feedback for dictation should favor content accuracy but include function accuracy. For example, if a learner writes survive in place of survey, that error would carry more weight than if a learner writes increase in place of increased. This activity can serve as both practice and assessment.

2. Dictogloss

In a dictogloss activity, learners create their own transcripts based on a short audio sample. Learners may listen to the sample multiple times until they feel confident in their transcription. Once learners have completed their individual transcripts, they compare what they have written with the actual transcript and mark the differences. They are directed to note differences in word boundaries, word endings, content word errors, and function word errors. Learners are then encouraged to reflect on their errors and any patterns that they notice. This can be done verbally in a class discussion or in writing.

3. Transcript Analysis

Transcript analysis activities work well as tools for predicting content and spoken language features. In this way, they can serve as a bridge between top-down and bottom-up listening strategies.

Predicting Content: Learners are given segments of a transcript and listen to the corresponding audio. After reading the transcript and listening to the audio, learners make predictions about what the speaker will say next based on the content and verbal cues. Students receive the next segment of the transcript and listen to the corresponding audio to check their predictions and make new predictions for the following segment. Learners continue this process through the entire lecture and discuss their strategies.

Predicting Features: Learners are introduced to the following pronunciation features of English and given examples of each: stress, thought groups, intonation, contractions, linking, reductions, and assimilation. They are then asked to analyze a partial transcript and make predictions about where they expect to hear each feature produced by the speaker. Learners listen, check their predictions, and discuss the activity.

4. Cloze Listening

Cloze listening tasks, in our experience, are the most commonly used activities for practicing and assessing bottom-up listening strategies. These tasks can be incredibly useful in assessing learner needs and raising learner awareness. In cloze listening tasks, learners are given a partial transcript with missing tokens (key words, numbers, etc.). Before the task, learners should make predictions about the content of each blank (part of speech, plurality, etc.). After the task, learners should reflect on the phonemic and semantic relationship between their answers and the target.

5. IPA/Pronunciation Practice

There is a notable lack of bottom-up listening resources in most EAP textbooks. One way to mitigate this absence is by using pronunciation activities as a tool for bottom-up listening practice. Many of the skills needed to perceive words in the speech stream are closely linked to the skills needed for accurate spoken production. Practicing these skills in tandem gives learners an opportunity to improve both their speaking and listening simultaneously. Information gap activities like minimal pair Battleshipgive learners a chance to practice their production and perception of challenging contrasts in a low-stakes context. Other examples of pronunciation-based listening activities include intensive and selective tasks in which learners are given minimally contrastive options and asked to determine which they hear in an audio sample.

In summary, the ability to process complex streams of speech remains a challenge for L2 learners at all levels, particularly for those who have academic interests in mind where accuracy is more important than gist. In our view, when language teachers skip some of these important bottom-up listening steps and instead focus on testing, not teaching listening, they overlook some of the crucial phases that are necessary to achieve more advanced levels of L2 listening comprehension. Our goal is to reconceptualize our L2 listening practice to be more holistic, so that the bottom-up skills are considered necessary building blocks for a more complicated, nuanced listening adventure.

References

Field, J. (2008). Listening in the language classroom. Cambridge, England: Cambridge University Press.

Goh, C. (2000). A cognitive perspective on language learners’ listening comprehension problems. System, 28, 55–75.

Jensen, E. D., & Vinther, T. (2003). Exact repetition as input enhancement in second language acquisition. Language Learning, 53, 373–428.

Matthews, J., & Cheng, J. (2015). Recognition of high frequency words from speech as a predictor of L2 listening comprehension. System, 52, 1–13.

Siegel, J. (2016). Listening vocabulary: Embracing forgotten aural features. RELC Journal, 47(3), 377–386.

Abigail J. Castle is an ESL lecturer at the University of Iowa. She holds an MA in linguistics with a TESOL focus. Her professional interests include bottom-up listening pedagogy, building active L2 vocabulary, and creating content-based modules for student success.