In a recent paper in the Journal of Curriculum Studies, Kirsti Klette — Distinguished Professor at the University of Oslo and one of the most important researchers working on classroom observation — asks a question that has haunted the teaching profession for decades: can we develop a shared language for describing what happens in classrooms?
The question is not new. Dan Lortie raised it in Schoolteacher in 1975, noting that teaching lacked the common technical vocabulary that other professions take for granted. A surgeon can describe a procedure and another surgeon, anywhere in the world, will understand precisely what is meant. A teacher describing "good questioning" could mean almost anything — from Bloom's taxonomy recall prompts to Socratic seminar facilitation — and the listener has no way to know which without watching the lesson.
Klette's argument is that the explosion of video-based classroom research and structured observation protocols over the last two decades has brought us closer to that shared language than at any point in the profession's history. Instruments like PLATO, CLASS, ICALT, and the various frameworks used in the QUINT studies have given researchers a vocabulary for describing teaching quality that works across national boundaries. When a Norwegian researcher and a German researcher watch the same lesson and code it using different instruments, they tend to agree on the broad dimensions of quality. The tools disagree on specifics — how to score, what unit to segment by, where to draw the boundary between "cognitive activation" and "instructional support" — but the underlying constructs converge.
Two things struck me.
The first is that Klette is describing a problem I recognise viscerally from the other side. The lack of shared language isn't just a research inconvenience. It is the reason that observation feedback in schools so often fails to produce change. When I was told as a student teacher that my "questioning could be stronger," the feedback was useless not because my observer was wrong but because neither of us had the vocabulary to specify what "stronger" meant. Was I asking too many closed questions? Was my wait time too short? Was I directing questions to volunteers rather than distributing them? Was I failing to build on pupil responses? Each of these is a different problem with a different solution, and "stronger questioning" distinguishes between none of them.
The observation taxonomy I built for Learning Lens was, in retrospect, an attempt to solve this exact problem — to give observers and teachers a shared vocabulary that is specific enough to be actionable. When an observer taps "Cold Call" or "Think-Pair-Share" or "Probing Follow-Up," they are making a claim about what they saw that another observer could verify and that the teacher could recognise. The qualifier sub-tags push further: not just "the teacher used retrieval practice" but "the teacher used low-stakes retrieval with individual whiteboards at the start of the lesson." That level of specificity is what turns observation from impression into evidence. And it is, I think, a practical contribution to the shared language that Klette is calling for — not from the research tradition, but from the classroom.
The second thing that struck me is the tension Klette identifies between generic and subject-specific frameworks. Some aspects of teaching quality — classroom management, the emotional climate, the pacing of a lesson — look similar regardless of whether the lesson is mathematics or history. Others — the quality of subject-specific explanations, the accuracy of representations, the cognitive demand of tasks — can only be assessed by someone who understands the discipline. This is a real problem, and one that any observation taxonomy has to confront honestly.
My taxonomy leans generic. It captures instructional practices that the research base suggests matter across subjects: questioning, checking for understanding, feedback, metacognitive scaffolding, guided practice. It does not attempt to assess whether a science teacher's model of photosynthesis is conceptually accurate, or whether a history teacher's source analysis task is appropriately challenging for the period being studied. That is a deliberate design choice, not an oversight. The observer using Learning Lens is typically a DHT or senior leader conducting learning visits across the school — they are not a subject specialist in every classroom they enter. The taxonomy gives them a language for what they can reliably observe. Subject-specific depth comes from the conversations that follow the observation, not from the capture instrument itself.
But Klette's work makes me think this is an area where the taxonomy could eventually grow. Not by attempting to code subject-specific quality — that way lies unreliable data — but by creating space for subject-specific qualifiers within the existing practice categories. The practice "Presentation of New Material" could, in principle, have qualifier sub-tags that are subject-sensitive: modelling a worked example in mathematics, close reading demonstration in English, source analysis scaffolding in history. The evidence layer remains generic. The descriptive vocabulary becomes richer. That is a direction, not a commitment — but it is one that Klette's argument makes more compelling.
The shared language of teaching that Klette envisions is not a single framework imposed from above. It is something more like a convergence — multiple traditions, arriving independently at similar constructs, gradually aligning their vocabulary through comparison and collaboration. The 2025 special issue of School Effectiveness and School Improvement that her work anchors demonstrates this convergence in action: six different observation frameworks applied to the same Nordic classrooms, with instructive patterns of agreement and disagreement.
Where practitioners fit in this conversation is less clear, and that is partly Klette's point — the shared language has been developed primarily by and for researchers. The challenge now is whether it can migrate into the operational vocabulary of the people who actually observe teaching every day: the deputy heads doing learning visits, the middle leaders running departmental observations, the mentors supporting early career teachers. If the shared language stays in the research literature, it remains an academic achievement. If it reaches the clipboard — or the iPad — it becomes a professional resource.
That migration is what I am trying to build.
References
Klette, K. (2023). 'Classroom Observation as a Means of Understanding Teaching Quality: Towards a Shared Language of Teaching?' Journal of Curriculum Studies, 55(1), 49–62.
Klette, K., Luoto, J.M. & Jentsch, A. (2025). 'Do Observation Tools Matter? Unpacking Shared and Distinct Patterns in Teaching Quality.' School Effectiveness and School Improvement, 36(3), 433–447.
Lortie, D. (1975). Schoolteacher: A Sociological Study. Chicago: University of Chicago Press.
Jamie Scobie writes from extensive experience in Scottish secondary education, including pastoral care, data for improvement, and school self-evaluation. This blog is an independent publication: he writes in a personal capacity as the creator of Learning Lens, writing about classroom observation, teaching evidence, and education policy. He speaks here only for himself and for Learning Lens, not for any employer or other organisation. He holds an MSt from Cambridge (Distinction) and a Masters from Stirling.