Category Archives: Literature notes

On Language Learning Strategies

A few days ago I attended a conference in Klagenfurt, which focussed on language learning strategies. Strategies are not really my field of expertise, so the conference was a good opportunity for me to learn more about the topic, and what I will try to do in this post is to summarise and synthesise what I learnt from the presentations that I attended.

This is, by necessity, a very partial report on the conference. Firstly, there is the perennial conference problem, that one cannot attend every presentation (is it just my impression, or are the presentations you want to attend always scheduled in the same time-slot?). And secondly, some very interesting presentations did not fit in well with the focus of this post, and probably deserve posts of their own (e.g., Stephen Brewer’s talk on the parallels between mastering linguistic and musical communication, or Zarina Markova’s reflexive account about doing strategy-related research).

In the paragraphs that follow, I will begin by defining language learning strategies; following that, I will discuss why they are important for language learning; and finally, I will present some implications for language education and research, which were brought up in the conference.

Sign writing 'University of Klagenfurt" in German, Slovenian and Sign Langauge
An inclusive welcome

What are language learning strategies?

Language learning strategies were defined by Carol Griffiths in her plenary as “actions chosen (either deliberately or automatically) for the purpose of learning or regulating the learning of language”.

This fairly broad definition encompasses quite a few behaviours (and non-behaviours), and in fact, there are references to dozens of strategies in the literature. The proliferation of strategies, and frameworks that have been used to impose some order on the chaos, can at times be somewhat confusing. In his plenary, Andrew Cohen helpfully suggested a broad taxonomy that classified strategies, according to:

  • Goal: Under this heading, a distinction was made between strategies that facilitate learning (e.g., identifying and recording new words), and strategies that facilitate performance (e.g., retrieval and communicative strategies)
  • Function: This heading was used to classify strategies under sub-headings such as cognitive, affective, meta-affective, social, and more.
  • Skill: This heading referred to whether the strategies focussed on listening, speaking, reading or writing.

An important point that was iterated several times in the conference was that the use of strategies is a situated experience. That is to say, both the choice of strategies, and their effect depend on a number of factors pertaining to the user and the communicative situation. Yu Tina Yang, who reported on a study with Native Speakers of English learning Chinese, concluded that the use of strategies was affected by the nature of the task, the learner’s background knowledge and individual differences (e.g. different needs and priorities). In similar vein, Griffiths suggested that the use of strategies depends on the task and on individual characteristics, such as motivation, beliefs about oneself and about the language learnt, and on the learner’s ability to work autonomously. On the other hand, she suggested that research on the effects of age, gender and learning style differences has been inconclusive. The comments made by Griffiths seem to be consistent with a study reported by Vee Harris and Michael Grenfell, who compared 120 students from inner-city and suburban schools in the UK, and found that while gender did not seem to have any effect into the frequency of strategy use, but motivation did.

One final observation about strategies, which was raised by Sarah Mercer, is that they seem to have a recursive effect on shaping the context out of which they emerged. What this means is that it is perhaps counterproductive to think of strategies and context as being conceptually distinct. Rather, the context appears to be in a constant state of flux, and this dynamism seems to be sustained, among others, by the learners’ use of strategies. This is important to bear in mind when thinking about how to research strategies, as we need to be sensitive both to the ways in which each language learning situation is unique, and to the ways in which the situation constantly changes in response to the use of strategies.

Why are strategies important?

Most talks in the conference were premised on the belief that the use of language learning strategies is associated, in one way or another, with successful language learning and use.

For instance, Carol Griffiths pointed out that the frequency of strategy use was associated with successful learning; that successful learners tend to use strategies more frequently than less successful ones; and that successful learners often know how to orchestrate their strategy repertoires to address specific language problems.

Christina Gkonou used empirical data to show that highly anxious learners were able to deploy strategies such as positive thinking, relaxation techniques and actively seeking support from their peers in order to regulate their anxiety in language learning situations.

Harris and Grenfell also showed evidence that learners who exhibited persistence were prepared to take risks and tolerated ambiguity in communication tended to make stronger gains in language learning, compared to those who gave up easily, were risk averse and for whom ambiguity triggered self-doubt.
Embed from Getty Images

What are the implications for language teaching?

Several speakers suggested ways in which the findings from strategy research could be used to inform language teaching practice. The most important implication, raised by Griffiths, is that strategies are teachable; therefore if we make an effort to teach strategies, and if we encourage their use in a classroom context, this could make language learning more effective.

Cohen also discussed this point at some length in his plenary, where he suggested making students more aware of the strategies that are available to them, conducting strategy-based instruction, and even formally assessing strategies in language learning settings. He also suggested that teachers can help learners select appropriate strategies, by making themselves aware of their learners’ preferred strategies, linguistic proficiency, cultural background, motivation and more – although that raises the question of how feasible such demands might be.

Perhaps more pragmatically, Harris and Grenfell noted that bilingual students seem to use language learning strategies more effectively, and suggested capitalising on their expertise, particularly in large multi-ethnic schools. They also noted that, according to their research, motivation among UK students seems to plummet at the age of 12-13, so it would be sensible for strategy-based interventions to focus on that age group.

Another concrete suggestion was made by Gkonou, who recommended that interventions inspired by cognitive psychology could be used to increase self-regulation and enhance the students’ emotional intelligence.

What are the implications for research?

Some of the most important points raised in the conference pertained to questions such as what research should be done on strategies, as well as ‘how’ and ‘why’ to research the topic. Perhaps the most frustrating of these was made by Peter Gu, who reported on a large-scale literature review of studies conducted in China, to conclude that while a large repertoire of strategies had been identified, the real world impact of this research has been limited, and researchers are a fault for this unfortunate situation. Among the problematic aspects of research that he identified was the fact that research is often published in journals that are inaccessible to teachers, and there has been little effort to connect such research to training materials and procedures. Although Gu was discussing Chinese educational research, his remarks were met with knowing nods among the audience. I, for one, was frustrated by a number of studies in the conference, including nation-wide surveys involving thousands of students, the cost and effort of which seemed to be quite disproportionate to their social utility.

At the end of the day…

I left the conference in Klagenfurt knowing much more about strategies than I did before I went there, and with a sense of respect for all the good work that is being done in the field. However, I could not help fighting back the impression that arguing too strongly for Strategy-Based Instruction could be placing one demand too many on teachers, who are already being held accountable for more aspects of learning than is reasonable. At minimum, it seems to me, teachers should not be expected to tease out the relevance and implications of studies that do not make an effort to critically examine how they might influence practice. To argue that there is a need for more, and better, use of language learning strategies in the language classroom is perhaps useful, but what we really need to be doing is exploring is which strategies work where, and why. These were questions that have not been consistently addressed in the literature or in the conference, but it seems that the language learning strategy research community is reorienting itself towards this direction, and I think that it is a very useful redefinition of focus.

How can complexity inform ELT? Diane Larsen-Freeman has some ideas

Yesterday, I blogged about this symposium that aims to explore the ‘connectivities’ of ELT, i.e, the ways in which ELT bridges languages, cultures and disciplinary boundaries. The more I think of the symposium topic, the more interesting it seems; but at the same time, I am becoming increasingly conscious that ELT theory has perhaps failed to provide a convincing account how to straddle the faultlines that run across the field.

And this reminded me of a video I watched recently, where Diane Larsen-Freeman, the co-author of Complex Systems and Applied Linguistics, discusses complexity theory, and the ways in which it can help us to think beyond the dichotomies that are somewhat too prevalent in ELT theory. I’ve embedded the video below and written up an overview, in the hope that it may be of help in case anyone is interested in preparing an abstract for the symposium (or if you just have a general interest in how complexity can inform ELT).

The video

Just to provide a brief summary, Larsen-Freeman starts by defining complex systems as systems that are made up of many interacting components, which produce higher-order phenomena through their interaction. She uses a bird flock as an example of a complex system: when you approach a feeding flock of birds, they all rise in unison like a single super-organism, and perhaps unexpectedly, they evidence coordination despite the fact that there is no central organiser.

She then goes on to point out that complexity can help to challenge dichotomies that have pervaded our thinking about English Language Teaching, such as process vs. product, form vs. meaning, etc. She argues that while these dichotomies often provide us with useful heuristics, but can unhelpfully obscure any connections between the phenomena they describe.

With this in mind, she suggests that complexity can offer a way to look into the connections between three dichotomous pairs that come up often in the ELT literature:

  • grammar process & product: Complexity helps us to understand how grammatical regularities originate in language use, rather than from the top-down imposition of formal rules. Emergent regularities then become sedimented into patterns, through a process of ‘grammaring’, and it is these patterns that then constrain future use.
  • lexis & grammar: This dichotomy has already been challenged by empirical work in corpus linguistics, which has raised awareness of lexico-grammatical phenomena. Lexico-grammar ranges from fixed phrases to semi-lexicalised patterns, and complexity theory can help to account for their use.
  • learners & learning: Larsen-Freeman cites evidence from emprical research including her own, which have suggested that while learners share a common learning process, they also go through unique developmental trajectories. In this case too, complexity can help us understand how the trajectories interrelate with shared learning processes.

Larsen-Freeman concludes her talk by suggesting some implications of these insights for English Language Teaching. For example, she suggests using iterative learning processes, that allow for creative repetition of language. She also recommends creating affordance-rich learning experiences, from which each learner might learn in different ways.

A word of caution

When engaging about a new, analytically powerful, and somewhat broad theory such as complexity, there is a danger of believing that this is the ‘correct’ way of thinking, and that previous approaches were ‘wrong’. I would argue that complexity is neither the ‘right’ way to thing about ELT, nor is it ‘wrong’; it is just one tool, out of many that make up our analytical toolbox. There are times when using complexity may not be the most appropriate analytical choice – to build on the toolkit metaphor, this might be like using a spanner to hammer a nail on a wall. There are other instances though, for which complexity is ideally suited to generate new insights, and I would argue that exploring the ‘connectivities’ across ELT theory one of them.

Featured image credit: Fractal flame (Wikipedia) CC BY-SA

Tissue rejection

It seems that it’s been a while since I wrote an ‘Asked and Answered’ post, but here’s an interesting question that found its way into my inbox:

What is meant by tissue rejection in language teaching methodology?

‘Tissue rejection’ is an evocative metaphor that was used by Adrian Holliday (1992) to describe what happens when a teaching method, which is known to work in a particular educational setting, is introduced into a different setting where it fails to catch on.

In Holliday’s early writings, a distinction was made between what he called BANA and TESEP models of instructed language learning. In BANA (British, Australasian and North American) settings, learning tends to take place in private language schools or language learning centres affiliated to universities, and there is often ‘relatively clear contract between institutes and mainly adult groups who come specifically to learn English’. By contrast, in the TESEP model, derived from the words Tertiary, Secondary and Primary, language learning does not usually have an instrumental objective (Holliday 1998: 12). Holliday has since moved on from this rigid binary distinction, but I will continue to use it in this post because it helps to more clearly illustrate the ’tissue rejection’ metaphor.

Each of these two models has evolved different methods, which are in line with local cultural expectations, learning materials and resources, classroom arrangements and so on. For instance, BANA education is often underpinned by what Holliday defined as the ‘learning group ideal’, which sets the conditions for ‘a process-oriented, task-based, inductive, collaborative, communicative English language teaching methodology’ (ibid: 54). TESEP educational settings, on the other hand, might privilege a more traditional, transmissive, form-focused approach to language learning, which is closer to the norms of mainstream education in those settings.

The problem, Holliday argues, is that there is a tendency for TESEP to be perceived as less-than-effective, and the remedy is thought to be the adoption of BANA models. This is, in a sense, similar to a situation where a patient undergoes an organ transplant. However, when such innovations take place, we do not (and cannot) replicate the entire BANA model in the new setting. Rather, what is transferred is a limited selection of methods, which often do not fit very comfortably in the new context where they are transplanted. The new method (the ’tissue’), which was effective in its original setting, then becomes a source of disruption in the new setting.

Tissue Rejection
Figure 1. Relative positions of the language learning ideal in different environments (from Holliday 1994: 105)

A common scenario of ’tissue rejection’ is when a language teacher tries to introduce pair-work or group-work activities in a class where learners have been accustomed to working individually, under their teachers’ guidance. In such a case, it’s likely that the learners start engaging in off-task behaviour, or become disruptive; fellow teachers might complain about the noise levels in the language class; and parents might question the language teacher’s professionalism. The key thing to remember in this case, is that the problem does not stem from the teachers’ classroom management skills. Rather, it is rooted on the mismatch between the culture from where the method originated, and the culture where it is being implemented. Returning to Holliday’s metaphor, it is similar to what happens when a patient’s immune system attacks an otherwise perfectly good organ that has been transplanted into to said patient’s body.

The ’tissue rejection’ image is a powerful metaphor that helps us to understand the social and cultural intricacies involved in teaching English worldwide. Although the premises on which it was originally grounded (i.e., the existence of two incompatible English Language Teaching models) have given way to more nuanced thinking, the spectre of tissue rejection is still relevant in at least two ways. Firstly, it highlights the need for language educators to be aware of, and sensitive to, the subtleties of local educational cultures. And secondly, it serves to remind us of the complex, and often unpredictable, ways in which different cultures of learning interact.

Featured image by Rob Ireton @ Flickr, CC BY 

Is there a talent for language learning?

Not too long ago, I was caffeinating at an academic conference, when I was joined by a few friends who introduced me to one of the student assistants. “This is Kevin”, they told me. “He’s a hyperpolyglot: he can speak with everyone in this room in their native language!” Upon learning that I was Greek, Kevin looked down in embarrassment and said that he didn’t know much Greek – and then his face suddenly lit up, and he embarked on an animated chant: “Greek lover, Greek lover, me triha san pullover [your chest hair’s like a jumper]” He followed up with a stream of profanity, which he delivered with the eloquence and creativity of a drill instructor. Much merriment ensued.

What prompted me to remember this story was this newspaper article which claims that the ability to learn languages is determined by our genetic endowment. The usual caveats against scientific journalism apply, but the article does raise an interesting question. If some individuals have exceptional language learning abilities, what exactly are these traits and how do they contribute to learning? Despite the claims made in the article, our level of understanding in genetics and neurolinguistics is just nascent, which means that we can’t answer this question with any degree of certainty. With this in mind, what I aim to do in this post is discuss what little we do know about language learning aptitude.

What is Language Learning Aptitude?

Language learning aptitude (LLA) is the ability to effortlessly learn languages, which can lead to high language learning outcomes. In the words of Peter Skehan (1989), it is a “pre-programmed autonomous language learning ability” (p. 33). The first thing to note about this definition is that aptitude and outcomes are distinct. That means that there are many individuals who possess high LLA, but have not had the opportunity or motivation to learn additional languages. Conversely, there are individuals whose impressive linguistic capabilities are the product of effort rather than predisposition. That said, however, LLA can predict learning outcomes to some extent.

Another important distinction to make is that LLA is not quite the same thing as general intelligence. A number of empirical studies have found that LLA has a low moderate correlation (0.4 to 0.45) with measures of intelligence, indicating that the two constructs are related but not identical (Gardner & Lambert 1972, Skehan 1982). It also seems that most individuals with outstanding linguistic performance have above-average intelligence, but not outstandingly so (Skehan 1998). Moreover, there have been several reports in the literature, about individuals who have mental retardation and yet are highly proficient in several languages (e.g., Smith & Tsimpli 1995).

Aptitude research was quite popular in the 1950s and 1960s, at which time a number of aptitude tests were developed, with a view to using them to inform language teaching. The best known of these was the Modern Language Aptitude Test (MLAT), which was developed by John Caroll and Stanley Sapon in 1959. This was originally used for screening applicants for the US Foreign Service Institute. Another version of the test, the MLAT-Elementary, was used for selecting and streaming children in language programmes in primary education. The Pimsleur Language Aptitude Battery, or PLAB, was another aptitude test used in secondary education settings. However, the decline of structural linguistics, on which these tests were based, and the arguably non-democratic ways in which LLA testing was being used, meant that aptitude testing fell into decline in more recent years.

Components of Language Learning Aptitude

So far I have discussed Language Learning Aptitude as a singular construct, and this makes sense in the interest of efficiency. However, LLA research has always viewed aptitude as a composite construct, which brings together several components. An influential model of language learning aptitude, which has been put forward by Peter Skehan (1998), distinguishes between three main components: phonemic coding ability, language analytic ability, and memory (Figure 1).

Figure 1. Components of Language Learning Aptitude (Skehan 1998: 217)
Figure 1. Components of Language Learning Aptitude (Skehan 1998: 217)
Phonemic Coding Ability

Perhaps counter-intuitively, different languages make different distinctions between sounds: e.g., Russians do not distinguish between /f/ and /θ/, and to a Greek /i:/ (as in “sheep”) and /ı/ (as in “ship”) sound the same. Roughly speaking, the Phonemic Coding Ability refers to the ability to tell certain sounds apart, even if one’s mother language does not discriminate between them. It appears that at the early stages of instruction, individuals with a highly developed Phonemic Coding Ability strongly outperform their peers, but this advantage wears off in the higher levels of linguistic proficiency. In other words, it seems to function as a threshold of sorts, which is necessary but not sufficient for high linguistic outcomes.

Language Analytic Ability

The second component of LLA is Language Analytic Ability. This is defined as “the capacity to infer rules of language and make linguistic generalizations or extrapolations” from linguistic input (Skehan 1998: 204). As seen in Figure 1, the Language Analytic Ability appears to have a linear and monotonic relation to linguistic proficiency. That is, the greater the Language Analytic Ability an individual possesses, the likelier it is that this individual will have a high linguistic proficiency. However, because this relation is linear, it seems that the Language Analytic Ability cannot adequately explain the outstanding linguistic proficiency of hyperpolyglots, unless additional factors are taken into account.


The missing factor is memory. At the lower levels of proficiency, it appears that the relation between memory and linguistic proficiency is linear. However, among individuals with high levels of linguistic proficiency, small differences in memory can create enormous advantages, i.e., its role increases disproportionately.


So far, in this post I’ve looked into what Language Learning Aptitude is, and into its components. Moving forward, I’d now like to ponder on some implications for ELT practice and research. Firstly, I think that there is enough empirical evidence to suggest that language learning aptitude correlates with learning outcomes. Furthering our understanding of the construct and its impact on learning, and raising teachers’ awareness of the topic both seem like useful goals – at least as useful as the study of other learner differences, like multiple intelligences, differentiated learning styles and others. In other words, we should not assume, on the basis of ideological assumptions, that all learners are identical and that aptitude can be disregarded.

Equally, we should not assume, on the basis of our continuing research, that differences in language learning aptitude mean that some individuals are a better fit for language education programmes than their peers. In other words, while a high language learning aptitude may provide some students with an advantage in learning, it cannot and should not be assumed that there are individuals who are not capable of learning languages.

More to read

Two of the more accessible introductions to language learning aptitude are Peter Skehan’s Cognitive Approach to Language Learning (1998, Chapters 8 & 9), on which I have drawn extensively for this post, and the section on aptitude in Rod Ellis’s Study of Second Language Acquisition (1994, pp. 494-499). For an overview of more recent research, you may want to consult the state-of-the-art article by Richard Sparks and Leonore Ganschow (2001), which covers literature spanning from 1991 to 2001. There are also a couple of interesting articles by Peter Robinson and Peter Skehan in a volume edited by Robinson (2002). John Carroll’s (1993) Human Cognitive Abilities, by one of the pioneers of aptitude research, offers some additional information, which is perhaps dated, but still quite useful.

There are a number of interesting case studies of individuals with exceptional LLA, some of whom were mentally challenged. Here’s a sampling:

  • Curtiss, S. (1977). Genie: A psycholinguistic study of a modern day ‘wild child’. New York: Academic Press.
  • Curtiss, S. (1988) ‘The special talent of grammar acquisition’. In L. Obler and D. Fein (eds.). The exceptional brain: Neuropsychology of talent and special abilities. New York: Guilford Press.
  • Smith, N. and Tsimpli, X. (1995). The mind of a savant: Language, learning and modularity. Oxford: Blackwell.

If you can get your hand on the original aptitude tests, they are actually quite fascinating (there is a copy of the MLAT at the University of Manchester library, complete with test papers, answer sheets, a cardboard marking mask and instructions). The references of the tests I’ve discussed in this post are below, in case you need to cite them:

  • Carroll, J.B. and Sapon, S. M. (1959) Modern Language Aptitude Test. New York: Psychological Corporation.
  • Pimsleur, P. (1966) Pimsleur Language Aptitude Battery (PLAB). New York: Harcourt, Brace, Jovanovich.

Featured image by Man vyi (Own work) [Public domain], via Wikimedia Commons

Revisiting the Target Language Democratisation narrative

Back in the early days of English Language Teaching, teachers were keen to make all learners speak the Queen’s English and viciously suppressed linguistic diversity. Thus, they perpetuated Imperial rule long after the Empire had imploded. Then, in the 1980s, Braj Kachru argued that some varieties of English had become established in post-colonial settings, and it was deeply unfair to exclude them from teaching. Adherents of the ‘Standard Language Ideology’, like Randolph Quirk, fought back in defence of that ‘single monochrome standard’ which ‘looked as good in speech as it did in writing’, but Kachru held his ground, and his view eventually prevailed. Then, in 2000, Jennifer Jenkins published The Phonology of English as an International Language, which legitimised ‘foreign’ accents, and introduced a new, even more democratic and even more inclusive way of thinking about English. This means that we can now teach English unencumbered by possible feelings of guilt. 

Or so the narrative goes… It is a neat, linear, plausible story, with clearly-defined heroes and villains, and a predictable happy end. It is a narrative found in many accounts of the historical development of ELT, and is often repeated in the English as a Lingua Franca literature. As a matter of fact, and in the interest of full disclosure, it is the story that I narrate in Chapter 2 of my own thesis.

So what’s wrong?

But it would seem that this account, which I will provisionally call the Target Language Democratisation narrative, may not be entirely consistent with the historical record. And this is just a fancy way to say that it could very well be sham.

Earlier today, I was reading The Linguistic Sciences and Language Teaching, a book written by M.A.K. Halliday, Angus McIntosh and Peter Strevens in 1964, and here’s what I came across in pp. 203-204 (emphasis added).

One of the most important changes that took place in the period between 1950 and 1960 was the acceptance that ‘to speak like an Englishman’ was not the obvious and only aim in teaching English to Overseas learners (as far as speaking ability was imparted at all): in this respect British acceptance of variations of accent in English used overseas has run ahead of American views. Although linguistic theory was applied to language teaching by American workers long before it became a widespread practice to do so in Britain, the linguistic consequences of social and political changes, and especially of newly won political independence, have been recognised and accepted by the British while many American language teachers remain troubled but unconvinced. Some American language teachers accept only American forms of English, but in the eyes of the British language-teaching profession one or other of the varieties of English that are growing up may in specific cases be of a kind more appropriate to the local educational systems than any form current in the British Isles. This acceptance is accorded to varieties of English such as those labelled ‘Educated Indian English’, ‘Educated West African English’.

If this account is true, it seems that Quirk’s objections notwithstanding, there was already a strong tradition in British ELT of accepting and tolerating diversity in ELT. And this trend began at least 30 years before Braj Kachru wrote about World Englishes. This passage also seems to cast doubt on narratives such as the one found in Phillipson’s (1992) Linguistic Imperialism, that claim that ELT had always been hostile to vernacular varieties of English, and that the ‘centre’ of the English speaking world (the Anglophone West, that is) had always been keen to impose its norms on the ‘periphery’. While a single source cannot constitute proof, the extract above is hard to reconcile with Phillipson’s version of ELT history.

Why should this matter?

These may seem to be trivial points, and I will readily concede that they make little substantial difference to actual teaching today. But still, I think that there are two somewhat important implications which should not be overlooked. Firstly, the discrepancy between the Target Language Democratisation narrative and the account found in this book from the early 1960s, calls into question the extent to which canonical versions of ELT history are grounded on historical data. This is disturbing question, which has also been posed by Richard Smith elsewhere, and which perhaps needs to be asked more persistently.

Secondly, such discrepancies raise the question of how and why ahistorical accounts of ELT come to be established. As critical professionals, we need to ask ourselves why the nuanced statements made by Halliday, McIntosh and Strevens have failed to register in the history of the profession. I can only speculate that our understandings of the past are always mediated by the perceptions of by those who came in prominence more recently, perceptions which may have been oversimplified in the context of the heated lingua-political debates of the 1980s. If the invectives sometimes found in the English as a Lingua Franca literature are any indication, some of us have been rather too eager to reduce alternative viewpoints to straw-men arguments. While this tendency may be problematic in the context of scholarly debate, raising such oversimplified accounts to historical status is even more troublesome.

And what about fixing this mess? To me, it sees that there’s a need to retell the history of our profession – this time drawing on the findings of a rigorous, data-driven research agenda. We need to go back to the original documents, the books and writings of previous times, and we need to interview those people who were involved in giving the profession its shape. there’s lots of work to be done, so if there’s anyone out there looking for a PhD topic, this might be a worthwhile one.

Featured image by Hindrik Sijens @ Flickr | CC BY-NC-SA 2.0