Design history, taught by students

Goal

Design history is conventionally delivered by lecture and assessed by recall, a pairing that produces students who can name the Bauhaus and cannot recognise its influence in front of them. This project inverts the arrangement. Each team becomes the class authority on a period, conducts primary research, and teaches it back, so that the discipline’s history is learned by being reconstructed rather than received.

The governing constraint is that the presentation must be designed in the visual language of its era. A viewer should be able to name the decade before reading a word. This is what converts a research assignment into a design assignment: the student cannot satisfy the criterion without having internalised the period’s typographic conventions, palette, compositional logic, and ornament. Historical understanding becomes a condition of the artefact rather than a claim about it.

Audience

Students in AVC 183, Digital Graphic Arts 1. Second-semester design students who have the software and are now accountable for judgment: what a period looked like, why it looked that way, and what it argued for.

Process

Six teams, two students each, dividing 1895 to 1989 into consecutive spans. The periods are assigned with a research hint rather than a reading list, since the object is to teach the student to find the material, not to find it for them.

The hints are the pedagogy in miniature. Each one names an omission rather than a fact · Vienna and Glasgow, not just Paris; Japan, not just the Atlantic · so the first thing a student learns about their period is that the canonical account of it is incomplete. The design-history survey is a construction, and they discover this by hitting its edges rather than by being told.

The instructional move here is peer teaching, and it is chosen for a specific reason rather than for variety. Preparing to teach material changes how it is encoded. A student who expects to be tested organises for retrieval; a student who expects to explain organises for coherence, and must therefore identify what is central, what is subordinate, and what they do not actually understand. The obligation to make a subject intelligible to someone else is the fastest way to discover the holes in one’s own account of it.

This is the protégé effect: learners who study in order to teach outperform learners who study to be examined, and they do so on the same assessment. The mechanism is not motivation. It is the cognitive work of restructuring material for an audience, which is precisely the work that recall-based assessment permits a student to skip.

Peer teaching also redistributes something the lecture format quietly withholds. When one instructor is the sole authority, students calibrate their understanding against a performance they cannot inspect. When a classmate teaches, the reasoning is visible, contestable, and near enough to the learner’s own that the gap feels closable. Vygotsky’s zone of proximal development names the same asymmetry: a peer who has just understood something is frequently a better guide to it than an expert who understood it twenty years ago and has forgotten what was difficult.

The structure follows from this. Six teams, six consecutive periods, and a requirement that every student respond to every presentation. No team is given the whole survey; each is given a sixth of it and made responsible for the rest of the class receiving that sixth. The survey exists only if the teaching works, which converts an individual deliverable into a collective one. A team that presents badly has not lost points; it has deprived twenty people of a decade.

The project then runs in three graded phases, so that research, design, and delivery are assessed as distinct competencies rather than collapsed into a single artefact.

Research. A minimum two-page draft of notes in a shared team document, with every source linked as it is gathered. Four to six organising categories, each carrying three to five strong exemplars and a written summary that will later become narration or on-screen text. The library is named explicitly, and so are the archives: Wikimedia Commons, the RISD Picture Collection, the Met Open Access Collection, the Library of Congress. Faculty librarians are among the most underused resources on any campus, and the assignment says so.

Design. Brainstorm, sketch, build. Tool is the student’s choice · Miro, Slides, Keynote, Prezi, NotebookLM as a research companion · with one prohibition that applies to all of them: no default templates. A template is a period’s visual language chosen by somebody else.

Presentation. Recorded with OBS, the departmental standard, published unlisted to the student’s own YouTube channel, and posted to Discord. Students comment on every classmate’s presentation, not a selection. They are each other’s audience, and a class of six periods only becomes a survey of design history if everyone watches all six.

On AI, the assignment takes a deliberate position on assessment validity rather than a rule about honesty. AI is permitted without restriction for research and orientation, and the assignment teaches it properly: role, task, format; then deep-research modes; then verification. It names Claude, Gemini, Copilot, and ChatGPT, notes which are free through the district, and states plainly that models hallucinate designer names, dates, and publication titles, which are precisely the facts this project turns on.

One line does the real work: the historical images must be real. AI-generated decoration is acceptable; an AI-generated poster is not. A student may use a model to understand De Stijl and may not use one to manufacture a De Stijl artefact. The evidence of learning here is located in the sourcing, the design judgment, and the live defence, none of which a model can supply. A generative model can produce a competent essay about Constructivism. It cannot locate a Rodchenko in an archive, credit it correctly, reproduce its compositional logic in a new artefact, and then explain the decision out loud to twenty classmates. The assignment is built so that the part a model can do is the part that was never being assessed.

Outcomes

By the end of the project a student will be able to:

• Research a design-history period using academic databases, library collections, and primary archives, and evaluate the credibility of each source.
• Organise a body of historical research into four to six defensible categories, whether movements, designers, themes, or spans within the era.
• Produce a designed presentation whose typography, colour, layout, and imagery reproduce the visual conventions of the period.
• Cite every image and source completely, including the artist, title, date, and provenance, and disclose the use of AI tools.
• Teach the period to peers in a recorded presentation, and respond substantively to every classmate’s work.

The project carries the college’s Visual Communication institutional learning outcome for analysis, critique and interpretation, which asks that a student interpret visual work and account for the choices behind it. A presentation designed in the visual language of its period cannot be produced without that reading, and cannot be defended to the class without articulating it. The outcome attaches to the presentation rubric.

An optional extra-credit track rebuilds the project as an interactive tool · a slideshow, a sorting game, a timeline, a quiz · produced by directing an AI rather than by writing code. Students are required to write a product requirements document before they open the model: what it does, who it is for, what the sections are, what it should look like, what happens when someone opens it. The PRD is the assignment; the tool is the by-product. Students who have only ever used a model conversationally discover that specification precedes generation, and that the quality of the output is bounded by the clarity of the intent. It is the single most transferable thing in the project, and it has nothing to do with design history.

The class example is a 1910 to 1920 slideshow with a drag-and-drop sorting game, built to demonstrate the standard rather than to be copied.

Student work

Status

In active use, taught each term on the student-as-teacher model. The optional AI build track, a product-requirements document first and the tool second, is the current addition. The student work shown is from real sections.