Describe-and-Draw Barrier Game

Two people sit with a barrier between them — a propped-up book, a cereal box — so neither can see the other’s paper. One is the describer, who has a simple picture (or builds one), and must describe it well enough that the listener can draw or build a matching copy using words alone. Then you lift the barrier and compare. The mismatch is the fun — and it teaches the child to say exactly enough for someone who can’t see what they see.

  1. Set up the barrier. Each player gets paper and the same crayons (or identical building pieces). Prop something up between you.
  2. The child describes; you draw what you hear — literally. If they say “a circle,” you don’t know it’s a big red circle in the corner; draw a guess and let the gap show.
  3. Resist helping. Don’t supply the missing words — let your honest mis-drawing reveal what was left out (“Hmm, I didn’t know where to put it”). The productive struggle is the whole mechanism.
  4. Lift the barrier and compare. Laugh at the differences, spot which words would have helped, and notice when the child checks in (“did you put it next to the tree?”).
  5. Swap roles. Being the listener — getting stuck on a vague description — is what teaches the child to describe better next time.

Variation: Build-don’t-draw — match LEGO/Duplo arrangements or pattern blocks instead. Draw-my-monster — the child invents a creature and describes it. Sticker scenes — place stickers by description. Back-to-back — skip the barrier and just sit back to back.

Requirements

  • Space: A small table or floor space for two; works anywhere two can sit
  • Surface: A flat surface for drawing or building
  • Materials: Two sheets of paper and two matching sets of crayons, OR two identical sets of building bricks/pattern blocks/stickers; a book or box for the barrier
  • Participants: 2 players — 1 describer + 1 listener (adult–child, or two children)
  • Supervision: Adult plays a role or lightly facilitates

Rationale & Objective

This is the classic task for building referential communication — describing something precisely enough that a listener who cannot see it can act correctly. Glucksberg, Krauss & Weisberg (1966) introduced the two-person barrier paradigm and found that 4–5-year-olds could not yet reliably describe novel referents to a hidden listener, establishing that informative, listener-adapted description is an emerging, trainable skill exactly at this age. And it is trainable: Bunce (1989) showed children trained in a barrier-game format made significant, lasting gains in referential speaking. The deep mechanism is perspective-taking — to describe well the child must suppress their own privileged view and represent what the listener does and doesn’t know; Sidera et al. (2018) found theory-of-mind skills predict the informative messages and clarification requests children produce in exactly this kind of collaborative task. The game makes that mental modelling concrete because the listener’s drawing instantly reveals where the description failed, teaching the rule “be as informative as required” without a single lecture. Honest framing — at 5 most children sit in the early bands (vague or single-attribute descriptions), and the value is the built-in feedback loop, not adult coaching; keep it collaborative (“let’s make them match!”) rather than a test, because the adult over-helping or scoring “right/wrong” removes the very struggle that does the teaching.

Progress Indicators

  • Early: points or says “it’s a thing”/“put it there”; names objects but no attributes; assumes the listener sees what they see; drawings barely match; sentences 2–4 words, often fragments
  • Developing: names objects plus one feature at a time (“a big dog”, “a red one”) but omits position, quantity, or relations and doesn’t sequence; the listener gets some elements right but placement wrong; sentences 4–6 words
  • Proficient: combines attributes and adds spatial/relational words in fuller sentences (“a small blue circle in the top corner, and under it a yellow star”); begins to sequence top-to-bottom; most elements reproduced; sentences 6–8+ words
  • Advanced: gives complete, unambiguous descriptions, checks in with the listener (“did you get the part next to the tree?”), answers clarification requests and reformulates when the listener is stuck; near-exact reproductions

Safety Notes

  • Keep it collaborative, never a test or a race — frame it as “let’s see if we can make them match!”; competitive or evaluative framing raises performance pressure and can make a shy child withdraw
  • Ordinary craft supervision only (markers, and small building pieces kept away from younger siblings — see the choking note on the Mystery Box game)
  • Sustained describing plus listener-modelling is cognitively demanding — keep rounds short (a few items) and stop while it’s still fun
  • Worth a speech-language view if, by 5, the child speaks only in 2–3 word fragments with no growth, can’t produce attribute words even with modelling, shows marked word-finding struggle, or shows no listener awareness across many sessions despite scaffolding

Hints

  • Playfulness: theme the target (build the same zoo, draw a silly monster); let the child design the picture for you to describe; use a real “secret” folder so the reveal feels like a surprise
  • Sustain interest: rotate materials — crayons, LEGO, pattern blocks, magnetic tiles, stickers; add a gentle rule like “no pointing” or “no peeking” to up the challenge
  • Common mistake: over-helping — finishing the child’s sentence or supplying the missing word kills the mechanism; let your honest mis-drawing surface the gap instead; no “correct/incorrect” scoring
  • Limited space: one pencil and paper each is enough; “describe the object behind your back” or back-to-back drawing works in a car or waiting room
  • Cross-domain: theory of mind and perspective-taking (representing the listener’s view); vocabulary (attribute, colour, shape words); spatial and math language (above, below, between, beside, corner, half); fine motor (drawing and placing); classification (grouping by feature)
  • Progression: place 2–3 large pre-drawn shapes on a grid → colour and size that force disambiguation (“the big red one, not the small one”) → free-build with blocks or stickers → describe a full picture scene with spatial relations → add a time limit or a no-pointing rule

Sources

  • Glucksberg, S., Krauss, R. M. & Weisberg, R. (1966). “Referential communication in nursery school children: method and some preliminary findings.” Journal of Experimental Child Psychology, 3(4), 333–342 — the foundational barrier-task paradigm
  • Bunce, B. H. (1989). “Using a barrier game format to improve children’s referential communication skills.” Journal of Speech and Hearing Disorders, 54(1), 33–43 — barrier games train referential speaking, with maintenance at follow-up
  • Sidera, F., Perpiñà, G., Serrano, J. & Rostan, C. (2018). “Why is theory of mind important for referential communication?” Current Psychology, 37(1), 82–97 — perspective-taking underlies informative description
  • Head Start ELOF — Language & Communication, Goal P-LC 5 (expresses self in increasingly long, detailed, and sophisticated ways) and Goal P-LC 6 (uses a wide variety of words for a variety of purposes)
  • UK EYFS — Communication & Language, Speaking ELG (express ideas in full sentences; offer explanations, with modelling and support)
  • ASHA — Communication Milestones: 4 to 5 Years (uses location words like behind, beside, and between; produces longer, more complex sentences)