The day I graduated from university, the "Con", I was on a plane from Sydney to Rome. Like many people from English speaking countries, I was purely monolingual.

I was not proudly monolingual. My father's incredulity at his eldest being unable to  count to ten in another language saw him generously supporting my immersion in European culture. The first stage of my Italian journey at the Università per stranieri in Perugia didn’t go well.

Having never studied a language, I didn’t understand what we were being taught. We turned the pages of a monolingual grammar book 20 hours a week in a class of 72 students, with a lecturer squawking at us day in day out. And this was of course the good ole days of grammar translation!

The second stage was much better. We turned the pages of the same book, starting at the beginning, in a class of c.15. In my teaching and training since, I have often recommended people work through the same book more than once.

Stage Three: Upon my return to Australia, I enrolled in a bachelor program at the University of New England  and studied Italian and German by correspondence.

It was much easier being a beginner in German because I understood what was being taught and what I had to do. My German studies is a topic for another post as it changed the direction of my life. In 1990, I found myself in Czechoslovakia and threw myself into Czech. I have recently spent four years in Uzbekistan and threw myself in Russian.

Since about 1985, I have not used Italian let alone studied it. I’m currently planning a getaway from the long harsh Czech winter. Sicily is my destination of choice, partly because I’d like to explore ChatGPT’s ability to help me revive my Italian.

Stage Four: 40 years later. … -ish

Hi ChatGPT, can you help me revive my Italian? I studied it many years ago.
Yes, of course. How do you want to proceed?
Let’s start with adverbs of frequency.
OK. Here’s a list.
Thanks. Can you put them into short simple sentences for me translate into English one at a time?
Sure. Here you go. 

When it uses words I don’t know, I ask.  When I notice a verb form that vaguely rings a bell, but I’m not sure of, I ask. It gives me a little grammar lesson – futures, irregulars, combinations with prepositions, subjunctives, etc.

Since becoming a ChatGPT paying customer, I wondered if it knew my name. I asked and it said no. I introduced myself and asked its name. It doesn’t have a name. I told it I’d call it Jax. I created Jax’s back story, which it promptly translated into Italian for me. I didn’t read it as I wanted to translate it myself, which I did, sentence by sentence.

I always ask in Italian and it corrects the wording of my questions and then Jax answers in Italian, sometimes using words and structures that I don’t know which I then ask about. 

It prefaces its feedback with:

• Non male, James!
• Quasi perfetto!
• Ottimo tentativo, James!
• Stai migliorando molto! 

I should really ask Jax to tone down his relentless praise. 

A few days ago, I suggested that perhaps there was too much new information and it was time to pause and revise. I wrote: Forse impararo troppo nuova informazione. Devo fare una pausa a lego tutti che abbiamo studiati ieri e oggi.

Jax replied with the corrected form of my attempt and an explanation of some of my mistakes: 

Capisco, James! Ecco una versione corretta della tua frase:
Forse sto imparando troppe nuove informazioni. Devo fare una pausa e leggere tutto quello che abbiamo studiato ieri e oggi. Ecco le correzioni:

• "impararo" → "sto imparando" (per il presente progressivo in italiano).
• "troppo nuova informazione" → "troppe nuove informazioni" (per accordare con "informazioni" al plurale).
• "a lego" → "e leggere" (per collegare congiuntamente con "fare una pausa").
• "tutti che abbiamo studiati" → "tutto quello che abbiamo studiato" (accordo e forma corretta del passato).

Prendersi una pausa per consolidare è sempre una buona idea!

So, you can see how Jax took my imperfect Italian, gave a corrected version and followed up with explanations of four key issues. Then he agreed that it was a good idea to pause for consolidation. It’s hard to imagine Jax ever disagreeing. 

I asked if the stress in agio was on “i”.

• Esatto, James! La parola "agio" ha l'accento sulla "i": à-GIO. It might seem tricky because of the soft -gio ending, but it's just two syllables!

I also had a tech issue and asked: I would like to know why everything we've done isn't appearing on my other computer. I thought that one account was one account. It suggested three approaches to the problem, one of which solved it. We did this English. 

We then continued. I told Jax that it was much easier to learn a second musical instrument than a first, and the situation was the same with learning a second foreign language. He corrected my attempt on Italian, then he congratulated me for gasping the parallel between music and languages. A tad condescending methinks, but I let it go! 

• Hai colto perfettamente il parallelismo tra musica e lingue!

I asked Jax if he played an instrument and he wrote a paragraph which launched a new discussion. He also wrote in 3rd person: Jax suona ogni tanto con amici. When challenged, he explained that he wanted to create a certain narrative distance, etc, 
 per creare una certa "distanza narrativa" e mantenere un tono descrittivo

When he then promised to write henceforth in first person, the system’s memory updated as it learned my preferences. 

And this is where we are at. We have covered much more stuff than I could report here. I’m having such a wonderful time with it that I’m wondering how long it will be before language teachers are supplanted altogether. In one of the dialogues in the phrasal verbs book I’m writing at the moment, one of the characters tells his teacher that she will never be replaced by AI, but he might be being disingenuous.

At the end of the new edition of Discovering English with VersaText, published yesterday, there is a list of 22 takeaways. These are the points that I hope I have instilled in the readers. They embrace teaching, learning, creativity, metalinguistics, metacognition (all things meta TBH), guided discovery, text, grammar, vocabulary, pronunciation, depth, respect and colour.

Here they are.

1. Text is the starting point.
2. The multiple affordances of texts and tasks.
3. The Six Pillars of Vocabulary – the properties of words.
4. All properties of words beget language learning tasks.
5. The task activates learning.
6. Relationships between words carry meaning.
7. Relationships between words and grammar carry meaning.
8. Disabuse students of the idea that vocabulary is random.
9. Reveal the visible patterns in language – patterns of normal usage.
10. Respect students' intelligence.
11. Students benefit from guided discovery.
12. Flatter students' intelligence by introducing terminology and its etymology.
13. Tell students the truth or lead them to discover it.
14. Train students in the art and science of observation and pattern hunting.
15. Take students on the journey from analysis to synthesis.
16. Prepare students to become lifelong learners.
17. Let them bask in the wonder of meaning creation.
18. Demystify English at all levels of the Hierarchy of Language.
19. Accept that students have to do the learning and teachers have to lead them to finding their own optimal paths.
20. Equip students with the information and skills they need to make guided discovery work for them.
21. Don't accept students' narrow view of language as a foundation for their wish list when asked what they would like to do or like to learn.
22. Never stop asking students, HDYK.

Originally, a Kindle only, it is now also a print book with white space for readers' answers, notes, comments. The new edition reflects all the updates we have made to the free, online software, and AI appears at pertinent moments.
The Kindle and book are available from Amazon.

​Should you be interested in all things VersaText, or even some of them, feel free to join the Facebook group too.

Language learners benefit greatly from example sentences, since it is an opportunity to learn language from language, my big thing. For this reason, I devoted a considerable amount of my teaching, training and writing to helping students gain the maximum benefit from illustrative sentences.

In the early 2000s, I attended my first Teaching and Language Corpora conference in Bertinoro, a beautiful hilltop town near Bologna, and presented my incipient formula for computationally selecting the most useful sentences from corpora to present to students. I programmed a tool that allocated the frequency of every word in a sentence and average it. Sentence length was also a criterion.

As mentioned in previous posts, the great English lexiocagrapher, Patrick Hanks was my colleague at this time and I asked him what criteria his kind used when selecting sentences to include in their dictionaries. He said there was no list.

I worked on this further and came up with a list of ten criteria that I discussed with Patrick and he added one more. I gave this list to Pavel Rychlý, who was developing Sketch Engine and his team used these criteria as a basis for their GDEX algorithm, i.e. good example sentences. It is now a standard part of SkELL and Sketch Engine. My criteria are listed on this 2006 webpage.

So, it’s a good thing that corpora can select illustrative sentences, but can students? And should they? In short, yes and yes. But then what? How does a learner know what they can learn from an illustrative sentence apart from it being a targeted piece of input which they might soak in, as they do from any input they are exposed to. The answer lies in knowing the properties of the target word that are necessary to shift it from active to passive use.

I am a strong advocate of the Collins COBUILD Advanced Learner’s Dictionary because it even presents its definitions in full sentences. Full sentence definitions are goldmines. From the sentence defintion, you can easily extract concept checking questions (CCQs). For example,

Collins: A wildcard is a symbol such as * or ? which is used in some computing commands or searches in order to represent any character or range of characters.

• What is a wildcard? A symbol
• What are some examples of wildcard? * or ?
• Where is it used? In computing commands
• What is its function? to represent any character …

Collins: An aphorism is a short witty sentence which expresses a general truth or comment.

• What is an aphorism? a short witty sentence.
• What does it do? it expresses a general truth or comment.

These sentences typically start with a hypernym, here symbol, which immediately limits what it is and is not. Their definitions progress with the target word’s features, functions, etc. Each of these is encapsulated in a phrase or clause in the sentence definition. They are the properties of the word. The Collins then provides example sentences in which the abstract properties are made concrete. If students know what they can learn from full sentence definitions, they can see how the meanings of words manifest in authentic sentences.

I’m writing a student workbook at the moment which will probably be called Discovering Phrasal Verbs, in which students are repeatedly tasked with finding example sentences in corpora. The book explains the importance of the semantics of the phrasal verb particles (prepositions and adverbs) and the importance of the subjects and objects of the verbs. These properties are the most important contributors to the meanings of the otherwise opaque, or at best translucent, phrasal verbs.

When you search corpora for a phrasal verb, the sheer volume of data can be overwhelming. Fortunately, SkELL uses GDEX, so the 40 sentences it presents are manageable. The other tool I recommend is CorpusMate because it is very fast, it enables searches with wildcards, and the cotext is colour-coded using the same colours for parts of speech as VersaText. The wildcard searches are necessary when the phrasal verb is separable, e.g. tear .* away, keep .* .* away.
​

AI is another source of illustrative sentences. In ChatGPT's own words, "The sentences generated by AI are original constructs, created using the language patterns learned during training." They are by definition inauthentic sentences, which means they were not motivated by any communicative impetus, hence they lack real-world contexts. These sentences often resemble those made up by textbook authors and test creators. It is reasonable to ask if the trade-off between authentic and inauthentic example sentences in terms of learnability is worth it. Do students really benefit more from authentic than inauthentic sentences?
​
Like all good questions in ELT, the answer starts with, it depends. My it depends revolves around what the students are tasked with. If the textbook provides made-up example sentences without any task other than perhaps read, read aloud, translate or memorise some sentences, the students will function at the bottom of Bloom's Taxonomy. Garbage in, garbage out.

​But if the tasks involve higher order thinking skills in which the students skim and scan multiple examples of authentic language in search of specific properties to which they have been alerted, they develop a better understanding of the properties of the target word, and ultimately a more sophisticated understanding of language per se emerges. Like all good citizen-scientists, students engaged in “extreme noticing” need systems to record their findings that will in turn deepen their conceptual grasp of the target language and prepare them to use it confidently.

It is well-known that guided discovery is not for everyone. I was a school music teacher in my 20s and one would occasionally hear, Never try to teach a pig to sing: it wastes your time and annoys the pig. This is yet another aphorism attributed to Mark Twain, but who knows?
​
Guided discovery demands a strong rationale, clear instructions, the right tools and an understanding that the students are going to benefit from the multiple affordances of the tasks. It is important that students are made aware of the multiplicity of these learning experiences in the process of acquiring words and their properties. No reflection, no connection.

I asked ChatGPT for ten passive sentences about the topic of connectionism. As usual, the sentences are bland and inauthentic. But this is one  was also problematic.

• Complex behaviors can emerge from the interaction of interconnected nodes in connectionist networks.

When I pointed out that this sentence has no passive voice, it apologized for the oversight and offered this instead.

• Complex behaviors can be emerged from the interaction of interconnected nodes in connectionist networks.

When I pointed out that emerged doesn't work in the passive in English, it told me that I was correct and offered me this.

• Complex behaviors can be observed to emerge from the interaction of interconnected nodes in connectionist networks.

When I pointed out that the infinitive form after observed doesn't work in English, it told me I was right again, and offered me this:

• Complex behaviors can be observed emerging from the interaction of interconnected nodes in connectionist networks.