decades of characterization of the novice programmer failure
phenomenon have not produced any improvements in learning,
nor pinpointed any credible cause. Predictably then, attempts at
curricular innovation, rooted in hunches and overwhelmingly
on the content side, have been ineffectual.
In contrast, the 2014 fMRI cognitive study of programmers
established that the brain makes sense of computer programs
in the regions of the brain long known to be associated with
language processing functions, not logic and not math. Although more research will be required to prove the case definitively, this physiological evidence puts a spotlight on an
aspect of programming instruction long taken for granted,
the Prescriptive Linguistics language model. Both the fMRI
study and the consistent anecdotal observations reported here
about the positive effects of memorization strategies on syntax acquisition constitute a compelling argument for investigating whether an alternate and frankly more promising approach—implicit language pedagogies informed by both SLA
theory and foreign language instructional principles—can enhance our instructional outcomes if scaled. On the flip side,
those wanting to devise new content or pedagogic approaches
to the introductory programming curriculum, but who ignore
the central cognitive role of language in programming, now
risk irrelevance. As Corder cautioned some fifty years ago, our
teaching will only succeed when it conforms to how the brains
of our students actually learn.
1. Amirtha, T. This Is Your Brain On Code, According To Functional MRI Imaging. Fast
Company, April 21, 2014; http://www.fastcompany.com/3029364/this-is-your-brain-
on-code-according-to-functional-mri-imaging. Accessed 2017 November 22.
2. Bailey, N., Madden, C. & Krashen, S. D. Is there a ‘natural sequence’ in adult second
language learning? Language Learning, 24, 2 (1974), 235–243. doi: 10.1111/j.1467-
3. Bennedsen, J. & Caspersen, M. E. Failure Rates in Introductory
Programming. Inroads - The SIGCSE Bulletin, 39, 2 (June 2007), 32–36.
4. Bransford, J., Sherwood, R., Vye, N. & Rieser, J. Teaching Thinking and Problem
Solving: Research Foundations. American Psychologist, 41, 10 (1986), 1078–1089.
doi: 10.1037/0003-066X. 41. 10.1078.
5. California CTC (Commission on Teacher Credentialing). Supplementary
Authorization Guideline Book (C TC, July 2016); https://www.ctc.ca.gov/docs/
2018 March 8.
6. Chomsky, C. The Acquisition of Syntax in Children from 5 to 10. (Cambridge, MA:
MI T Press, 1969).
7. College Board. AP Data: Program Summary Report 2017. (College Board, 2017);
Summary-Report-2017.pdf. Accessed 2017 November 22.
8. Corina, D. P., Lawyer, L. A., Hauser, P. & Hirshorn, E.. Lexical Processing in Deaf
Readers: An fMRI Investigation of Reading Proficiency. PLOS ONE, 8, 1 (2013), 1–10.
9. Corder, S. P. The Significance of Learner’s Errors. International Review of Applied
Linguistics in Language Teaching, 5, 4 (1967), 161–170.
10. CSTA Standards Task Force. CSTA K- 12 Computer Science Standards. (New York:
CSTA/ACM, 2011); http://c.ymcdn.com/sites/www.csteachers.org/resource/resmgr/
Docs/Standards/CSTA_K- 12_CSS.pdf. Accessed 2017 November 22.
11. Dulay, H. C. & Burt, M. K. Should We Teach Children Syntax? Language Learning, 23,
2 (1973), 245–258. doi: 10.1111/j.1467-1770.1973.tb00659.x.
12. Eklund, A., Nichols, T. & Knutsson, H. Cluster failure: Why fMRI inferences for spatial
extent have inflated false-positive rates. Proceedings of the National Academy of
Sciences, 113, 28 (2016), 7900–7905.
13. Floyd, B., Santander, T. & Weimer, W. Decoding the representation of code in the
brain: An fMRI study of code review and expertise. In ICSE ‘ 17 Proceedings of the
39th International Conference on Software Engineering, (2017), 175–86. doi: 10.1109/
14. Hamrick, Phillip. Ph. D. Dissertation, Georgetown University, Washington D.C.
Development of Conscious Knowledge During Early Incidental Learning Of L2
Syntax. (ProQuest, LLC, 2013), 3642262.
credentialed teacher knows that the ability to solve problems
is—and algorithms themselves are—highly domain-specific;
and that good problem-solvers draw upon prior experience and
knowledge of specific domains [ 4, 34]. Even so, problem-solving
is a process that is still poorly understood. 22
Finally, note that the approach described in this section will
of necessity require that instructors allocate considerably more
time to students practicing a unit’s concepts and sub-concepts
than the current instructional model provides. The instructional tension between breadth (coverage) and depth (detail)
is nothing new. Introductory programming language courses have traditionally opted for breadth given the very limited
amount of time they have allotted—or more precisely, that they
have self-imposed upon their programs of study. Such a limited
time frame could never work in a foreign language curriculum,
where four semesters are typically allocated for students to acquire proficiency in the fundamental workings of a language.
Foreign language curricula probe every topic in depth, because
the breadth of the curriculum can be adequately covered over
the two years budgeted for the program’s foundational sequence of courses. Unfortunately, there is no way to reconcile
the existing programming language model and course structure
with both breadth and depth learning. Something will have to
give, and the only resource available is time.
In his seminal paper on the errors made by second language
learners, S.P. Corder asserted that errors provided evidence
of—and insights into—the process by which learners construct
and refine hypotheses about the underlying grammar of the
language data they hear. He wrote.
We have been reminded recently of Von Humboldt’s
statement that we cannot really teach language, we can
only create conditions in which it will develop sponta-
neously in the mind in its own way. We shall never im-
prove our ability to create such favourable conditions un-
til we learn more about the way a learner learns and what
his built-in syllabus is. [ 9]
Yet, it is exactly the way novice programmers learn that has
continued to remain a mystery, obscuring how improvements
to teaching might be achieved.
No one denies that the current introductory programming
pedagogic model leads to less than favorable learning outcomes—particularly at the secondary level. The model undoubtedly contributes to ongoing low secondary enrollments
and the worst demographic inequities of any subject area. Four
22 From a psychological vantage point, problem-solving is a complex phenomenon,
described by Gestalt theorists with notions like “restructuring,” “insight,” and
“entrenchment;” and by cognitivism with a reliance on domain knowledge and heuristics.
In all of these, although conditions that facilitate the crucial moments of insight can be
listed, there are no satisfactory explanations for how such insights arise. The incubation
phenomenon—setting aside a problem after being unable to find a solution, with a
solution later popping into one’s mind (like a forgotten detail that one remembers after
the fact)—argues that problem-solving may be a largely subconscious process.