AI, AI, AI, AI! (AI takes over creating personalized, haptic pronunciation teaching homework!)

 A core component of the English Accent and Pronunciation Improvement Course is the weekly embodied oral reading. Each week students are trained in a "Movement, tone, tempo and. touch technique" (MT4) that is then practiced in the oral reading in various ways. Part of the process is the student applying the MT4 to a ~300 word text. The personalized text will be generated by AI (I use Perplexity currently, but any Chat-GPT arrived at a similar "solution.") Here is a simple example. (The basic prompt, fill-in-the-blank format is provided for students each week. In this case, the career discipline or context area, accounting, and the proficiency level and L1 of students are filled in.) 

• Generate a word list that includes 20 technical terms used in accounting that can be difficult to pronounce for advanced adult learners of English whose native language is Korean. Include the phonetic transcription of each along with the word, itself. 
• Create an engaging, 300 word story with as much dialogue as possible, using those 20 words, where the basic theme is where an accountant is probably stealing from one of the firm's clients.

WHAT FOLLOWS IS WHAT PERPLEXITY PRODUCED: 

-------------------------------------------------------------------------------

The Case of the Missing Revenue

At the prestigious firm of Ledger & Logic, the morning began as usual: the aroma of coffee mingled with the faint scent of freshly printed balance sheets. But today, something was off. Ms. Park, the head accountant, was hunched over her desk, eyes darting between the income statement and the cash flow report, her fingers nervously tapping the edge of the fiscal calendar.

“Strange variance in the revenue again,” she muttered, highlighting numbers with the intensity of a forensic auditor. The asset manager, Mr. Kim, peeked over her shoulder. “Did you check the subsidiary accounts? Last month’s reconciliation didn’t quite add up.”

Ms. Park nodded, flipping through a stack of inventory reports. “I did. But there’s an expense entry for ‘Mystery Amortization’—and I can’t find any supporting documents.” She frowned, recalling the audit last quarter, when a similar anomaly appeared under ‘depreciation’ for office plants.

As the team gathered for their weekly consolidation meeting, the boss announced, “We need to address the liability on our client’s books. The equity position looks weaker every month.” The room fell silent. Even the intern, who had just learned the difference between accrual and actual cash, sensed trouble.’

Suddenly, Mr. Kim spoke up. “I noticed the same thing in the reconciliation for our biggest client. The audit trail leads to a fiscal black hole—right into Mr. Cho’s office.” All eyes turned to Mr. Cho, who was busy pretending to do a subsidiary inventory count.

With a dramatic flourish, Ms. Park produced the incriminating balance sheet. “It’s clear: someone is siphoning off assets and hiding them under bogus expense and depreciation entries. The only question is—how long has this been going on?”

The office buzzed with whispers. In accounting, numbers never lie—but sometimes, accountants do.

The word list: 

Here is a list of 20 technical accounting terms that can be challenging for advanced adult English learners whose native language is Korean. Each term includes its standard IPA phonetic transcription, highlighting sounds that are often difficult for Korean speakers (such as /r/, /l/, /v/, /f/, /θ/, /ð/, /z/, and certain vowels).

Term IPA Pronunciation

• accounting /əˈkaʊn.tɪŋ/
• asset /ˈæ.sɛt/
• liability /ˌlaɪ.əˈbɪl.ə.ti/
• equity /ˈɛk.wɪ.ti/
• revenue /ˈrɛv.əˌnuː/
• expense /ɪkˈspɛns/
• depreciation /dɪˌpriː.ʃiˈeɪ.ʃən/
• amortization /əˌmɔːr.taɪˈzeɪ.ʃən/
• accrual /əˈkruː.əl/
• reconciliation /rɪˌkɒn.sɪl.iˈeɪ.ʃən/
• inventory /ˈɪn.vənˌtɔːr.i/
• audit /ˈɔː.dɪt/
• balance sheet /ˈbæ.ləns ʃiːt/
• income statement /ˈɪn.kʌm ˈsteɪt.mənt/
• cash flow /kæʃ floʊ/
• fiscal /ˈfɪs.kəl/
• variance /ˈveə.ri.əns/
• subsidiary /səbˈsɪd.iˌɛr.i/
• consolidation /kənˌsɒl.ɪˈdeɪ.ʃən/
• reconciliation /rɪˌkɒn.sɪl.iˈeɪ.ʃən/

Notes on Pronunciation Challenges:

• /r/ and /l/ distinctions (liability, reconciliation, accrual, balance)
• /v/ and /b/ (revenue, variance)
• /f/ and /p/ (fiscal, profit)
• /θ/ and /ð/ (method, although not in this list, is common in accounting)
• Vowel differences, especially /æ/, /ɪ/, /ʌ/, /ə/
• Stress patterns in multisyllabic words (depreciation, amortization, consolidation)
• These terms are essential in accounting and are frequently sources of pronunciation errors for Korean speakers due to the differences in sound inventories and syllable structure between English and Korean.

------------------------------------------------------------------------------

Not bad, eh! For this lesson, the MT4 is breaking up the sentences into rhythm groups, identifying the word with the most emphasis and then gesturing (jesturing!) on that word with the left hand tapping on the right deltoid muscle, termed the Syllable Butterfly. (You can see the training for that in this lesson from the Spring 2025 course.) 

Back to the future of (haptic) pronunciation teaching: beginning June 23rd!

The 2025 summer English Accent and Pronuncition Improvement Course (EAPIC) begins next month on the 23rd. (The Spring course materials and videos are still available here.) I asked Perplexity about it, since it will figure in very prominently in all lessons of the new course:

"The English Accent and Pronunciation Improvement Course (EAPIC), built on Acton’s haptic pronunciation teaching, is notably effective for English learners of all ages, abilities, and backgrounds due to its unique integration of movement, touch, and embodied practice. Unlike traditional pronunciation methods that rely heavily on auditory and visual cues, haptic pronunciation teaching (HaPT) systematically incorporates gesture and touch, creating a multisensory learning experience that enhances memory, recall, and confidence in spoken English.

This approach draws inspiration from voice training, embodied oral reading, and kinesthetic learning, allowing learners to physically anchor sounds and patterns in their bodies. Such “haptic anchoring” supports deeper integration of new pronunciation habits, much like muscle memory in sports or music training.The method is adaptable, making it accessible to learners regardless of age, native language, or prior phonetic knowledge. Regular, structured practice—both in and out of class—ensures that improvements are not only rapid but also sustainable, while the engaging, movement-based activities boost motivation and make learning enjoyable.

Ultimately, Acton’s haptic method empowers learners to internalize English pronunciation in a way that is memorable, meaningful, and effective across diverse learner profiles."

Could not have said it better myself . . . (In fact most of those terms and phrases are extracted directly from the blog and publications I have been involved with!) The embodied oral reading of the course will be all AI generated--by the students themselves--to provide the optimal context and personal connection between the genre, vocabulary and proficiency level of the learner and the practice materials. 

The weekly, 30-minute course training videos are free, on Youtube and Vimeo. Enrolment in the 60-minute live feedback sessions is optional ($250 USD). 

Join us next month! 

The future of pronunciation teaching: Extinction

Have just begun working with the new AI bot "perplexity." Asked it about embodiment in language learning. Although its primary source isn't particularly impressive--to put it mildly, its statement of the problem is surprisingly current/relevant. 

"Embodiment presents a significant challenge in learning a foreign language because it highlights the complex interplay between cognitive, sensory, and contextual factors. The theory of embodied cognition suggests that language learning is deeply tied to physical and sensory experiences. For instance, understanding words often involves simulating real-world experiences associated with those words, such as imagining the act of cooking when hearing the term "cooking".  . . Furthermore, studies show that second languages learned in formal settings tend to exhibit reduced embodiment compared to native languages. . . " 

And, according to the recent book, The Extinction of Experience: Being Human in a Disembodied World, by Christine Rosen, it can only get worse! (Also, check out her recent appearance on Uncommon Knowledge with Peter Robinson. ) Rosen does a remarkable job of exploring the impact of technology upon our connection to reality and how that is radically changing how we are able to relate to and draw from our "experiences." The first line of the promo introduces the theme well: "A reflective, original invitation to recover and cultivate the human experiences that have atrophied in our virtual world."

The impact on language teaching, and pronunciation teaching specifically, is only beginning to be understood. In the following series of blog posts I'm going to use Rosen's book as a point of departure. I begin with the guess (or maybe fantasy) that at least the haptic pronunciation teaching model can actually be a powerful and dynamic part of the way back. How ironic (or sweet!) that pronunciation work may make a significant comeback as being a key purveyor of embodiment in the classroom and elsewhere. 

Keep in touch! 

Sources

(from "perplexity" AI) The Embodiment of Language and Conceptual Knowledge https://www.mindbrained.org/2020/03/the-embodiment-of-language-and-conceptual-knowledge/

Rosen C. (2024). The extinction of experience: being human in a disembodied world. New York: Norton & Company, 

60 days to better pronunciation (and making a habit of it!)

 I'm reminded of the old saw: Even a blind squirrel finds a nut sometimes . . . In the midst of this meta analysis-lite of what it takes to establish a habit in Australia (summarized by Neuroscience news), there was a little gem or two from the research that supports what I have been saying to students for decades: practice pronunciation in the morning! There is no real explanation as to why that should be the case, just the numbers, but it makes perfect sense for several reasons. 

The more interesting conclusion of the study is that the current "consensus" as to how long it takes to establish a habit, that is about 30 days, does not apply to a wide range of behaviors Australians need/want to change. Some, it turns out can take up a full year to lock in. The research suggests that the minimum time is closer to 60 days, depending on any number of parameters. In the sports/fitness business opinions range from 1 to 3 months. (The popular "book" on changing habits is probably Atomic Habits: An Easy and Proven Way to Build Good Habits and Break Bad Ones by Clear, a great source on how to do it.)

What the researchers and Clear agree on, however, is that the process is not mysterious but generally grounded in regular, disciplined, systematic practice. In pronunciation work at least the engagement must also be fundamentally embodied, as well. . . . In other words, to be successful you need to not only stay with it for a couple of months, at least, but also need a scaffolded goal-focused plan to get there with. 

Next week, beginning on the 6th, is the English Accent and Pronunciation Improvement (EAPIC) course, pronounced: EPIC. It lasts 9 weeks, 63 days, just over mark for establishing a habit, according to the researchers. What it does is train the student's body (and pre-frontal cortex) in how to keep on improving, centered on what we call "Embodied oral reading:" (EOR), that is daily oral reading, in private, accompanied by a set of approximately 24 gestures from the course. The 8 training videos will be free on Youtube. They are enough for you if you are really disciplined and commit to practicing everyday, for about 20 minutes, on your own. If you need weekly feedback on your work, you can sign up for the course ($200 USD), too. The feedback sessions are always the following Wednesday on Zoom at 8 EST. 

If interested, connect w/me at: wracton@gmail.com

Original source:

Ben Singh, Andrew Murphy, Carol Maher, Ashleigh E. Smith. Time to Form a Habit: A Systematic Review and Meta-Analysis of Health Behaviour Habit Formation and Its Determinants. Healthcare, 2024; 12 (23): 2488 DOI: 10.3390/healthcare12232488

Guilt by reason of "accentedness" (and what should be done about it!)

Interesting study out of the UK: “Stereotyped accent judgements in forensic contexts: listener perceptions of social traits and types of behaviour” by Alice Paver et al., summarized by Neuroscience.com, as: "Do accents influence guilt perceptions?"  (I might also add that accents influence getting work!) It raises so many issues that I'd recommend that you read the full article yourself. The summary is not sufficient but is certainly provocative!!! Here is the Neuroscience,com summary: 

"Researchers analyzed responses from 180 participants who rated voices from 10 UK accents on social traits and likelihood of certain behaviors, including crimes. . . ." Leaving aside some obvious potential shortcomings of the design--some of which are acknowledged by the reseachers, such as using male speakers only and a design that sets up the focus on the "bias" before hearing the samples--the conclusions are . . . striking, to say the least:

"Accents influence perceptions of guilt, with those judged as “lower status” being considered more likely to commit crimes."

Now assuming that the results hold up later with

• An acceptable definition of what constitutes and accent
• Replication involving the other gender(s) and 
• Possibly a different general elicitation format, and that 
• The working class dialects do come with features that could  undermine the credibility or "hiring potential"--an intuition strongly confirmed or established in research over the decades . . . 

What should our approach be in the classroom in working with students who come to us with "working class" dialects who are aiming at white collar careers, for example? 

First, one of the other possibly relevant findings was that nonnative accents of the dialects tended to be seen as more trustworthy than the native speakers in the samples. Although it was not reported specifically which nonnative accents carried that "advantage," that sounds like good news for those who'd rather not get into accent work in the first place. Maybe. The distinction between "accent" and "pronunciation" that I give students is something like:

•  If when speaking slowly, your listeners have trouble understanding you, you need pronunciation work. Basic rhythm, stress and intonation instruction is key at that level. 
• If when speaking quickly and maybe under some stress, your listeners have trouble understanding you, you need accent work. Requires attention to better, more accurate production of key/professional terms and dialect features. pacing and voice quality settings. (May even include breath, posture and self-moitoring training.)

So, if your students come to you in a position where they have "absorbed" the features of a less prestigious, disadvantageous dialect and they are preparing for job interviews, f you can't help them at the accent-level, you may need work . . . or you may be doing so already and not know it! If you do need to upgrade your accent work toolkit, join us for the next haptic course next month!

Keep in touch!

Bill

Credit:
Clker.com

Original Research: Open access.
“Stereotyped accent judgements in forensic contexts: listener perceptions of social traits and types of behaviour” by Alice Paver et al. Frontiers in Communication

English accent and pronunciation improvement course LITE (EAPIC!)

Course begins with recorded introduction on February 6th!

In about two months make good progress in improving your English accent or pronunciation and be trained in how to keep on improving after the course. This is an 8-week, shorter more compact version of the regular English Accent and Pronunciation Course. (See the description of the full 12-week course here.) 

• The recorded, 30-minute weekly training lessons are free, available on Thursdays on Youtube.
• The next week on Wednesday at 8 p.m. EST there is a live  Zoom feedback session on that lesson for students who are enrolled in the course (for $200 USD.)
• The feedback sessions are recorded, so you can always access them later if you are enrolled. 
• At least be sure to watch the free Thursday training videos!!!

To really benefit from the course, you should do all the assigned homework for enrolled students, about 30 minutes per day. This is also a good course for an instructor who wants an introduction to haptic pronunciation teaching. 

The course officially begins Thursday February 6th, but you can join it anytime and still get caught up!  To enrol, just pay the $200 for the weekly Zoom feedback lessons and the homework assignments.

 All lessons are recorded and available later in day after the lesson. 

 For more information or to sign up, contact me, william.acton@twu.ca.

Keep in touch!

Bill

Bill Acton has been teaching accent and pronunciation improvement for about 50 years! This course, Version 7.0 of KINETIK system, contains many new techniques based on research in pronunciation teaching and neuroscience. Bill is now living in Chattanooga, Tennessee.