doodlchat

WHAT THE EXPERIMENTS FOUND

Draw it,
don't say it.

A drawing game is a conversation in pictures.

Two people at a wooden table playing a drawing game, one sketching simple shapes on paper while the other leans in to guess
AI-generated image, not a photograph of a real person or place.

FEWER LINES same meaning

76% Correct guesses the first time an object was drawn

When researchers asked pairs to communicate only by drawing, round after round, the pictures changed: busy sketches turned into a few quick lines that the partners understood instantly and outsiders often could not.

THE SHORT ANSWER

Pictures shrink
when partners talk back.

Psychologists use drawing games as a small model of how signs and words get started.
Three findings come up again and again.

ROUND ONE

Draw everything

In one study the first drawing of theatre had a stage, curtains, actors and an audience. Early signs work because they look like the thing.

Follow the rounds

LATER ROUNDS

Fewer strokes

With feedback from a partner, the drawings got simpler. By game six, theatre was a line and two circles, and the pair still understood it.

See the numbers

OUTSIDERS

Hard to read cold

People who had not taken part found the late, simple signs harder to identify than the early detailed ones. The meaning lives partly in the shared history.

Play it yourself

FOLLOW A DRAWING

What happens
to the picture?

Pick a moment in a repeated drawing game. The bar shows how far the drawing has moved from a detailed picture towards a quick shared symbol.

Round check

Choose a moment
The first roundDetailed picture

Drawers put in lots of detail so the guesser can recognise the thing by how it looks. In one chair study, guessers already picked the right one 76% of the time against 25% by chance.

Same pair, after several roundsQuick symbol

With a partner who answers back, drawings use fewer strokes and become simpler. The pair get faster and more accurate at the same time.

Drawing again with no partnerNo shrinking

When people drew the same items repeatedly without interacting, the drawings did not simplify; they became more complex.

A stranger sees the late drawingsHarder to read

Simple late signs were harder for non-players to identify than the detailed early ones. Viewers who followed one pair's drawings in order improved more over a session than those shown a shuffled mix.

A group of eight swapping partnersShared by all

In groups that rotated partners, one version of a sign spread until nearly everyone used it, like a word catching on.

Sources: Fay and Ellison, PLOS ONE (2013), describing Garrod et al. (2007) and Fay et al. (2010); Hawkins, Sano, Goodman and Fan (2021 preprint).

Summarised from published experiments; results describe groups of players in a lab, not any one game.

TRY THE EXPERIMENT AT HOME

Six rounds,
one stack of paper.

You can rerun the basic set-up with a friend. The lab versions banned speech, letters and numbers, so all the meaning had to go into the drawing.

A hand drawing a quick, simple outline of a chair in a few pencil strokes, with a real wooden chair blurred behind
AI-generated image, not a photograph of a real person or place.
  1. Make a list of confusable items

    Pick items that are easy to mix up, such as theatre, cinema and concert, so a quick sketch has to be precise. The guesser sees the whole list.

  2. Draw without words

    No speaking, letters or numbers. The guesser points to an answer and the drawer sees if it was right.

  3. Swap roles and repeat

    Go through the same list again, swapping drawer and guesser, for about six rounds.

  4. Show a stranger the last round

    Put your final drawings in front of someone who did not play, and see how many they can work out.

TWO NUMBERS WORTH KNOWING

Partners read
what outsiders miss.

From a drawing game where one person sketched one of four similar chairs
and a partner had to pick it out.

76%

Correct guesses the first time an object was drawn

Chance was 25%, as the guesser chose from four objects. Over repeated rounds, sketchers used fewer strokes and pairs became faster and more accurate.

FIRST DRAWING · HAWKINS ET AL.

88 vs 75%

Accuracy: real partners vs later viewers

Partners in the game were right 88% of the time overall. Later viewers who saw the same pair's drawings in sequence, without taking part, were right 75% of the time.

INSIDERS VS OUTSIDERS · HAWKINS ET AL.

Hawkins and colleagues also found the drawings did not drift into arbitrary scribbles: they kept the visual features that best set the target apart from the others.

WHY RESEARCHERS CARE

A small model
of how words begin.

Researchers in this field call the shift from look-alike pictures to arbitrary signs symbolisation. Because spoken language leaves no trace of its origins, drawing games give a way to watch something similar happen in an hour.

A 2018 study by Justin Sulik found that guessing early, detailed signs relied more on the creative, insight-style thinking used for puzzles, and that this reliance fell as signs became familiar. Familiarity, not how complex a drawing looked, predicted the change.

A detailed pencil sketch of a small theatre stage with curtains and a row of audience heads on a sheet of paper
AI-generated image, not a photograph of a real person or place.

Questions

Questions people ask

Why do drawings get simpler in a repeated game?

Drawing takes effort, and a partner who has already seen the full picture needs less to recognise it. In the experiments, simplification happened when partners could interact and give feedback.

Does drawing the same thing over and over make it simpler on its own?

Not in the studies described by Fay and Ellison. When people drew repeatedly without interacting with a partner, the drawings did not simplify and actually became more complex.

Why can't other people read our late drawings?

Much of the meaning comes from your shared history with your partner. Experiments found that non-players had more trouble identifying the late, simple signs than the early detailed ones.

Did the drawings become completely random marks?

In the chair study by Hawkins and colleagues, no. The sketches got sparser but kept the features that best told the target apart from the similar objects beside it.

What happens in a bigger group?

In 8-person groups where players rotated through every partner, one version of a sign tended to spread until almost everyone used the same refined drawing.

Where the facts come from

Updated

  1. Fay N, Ellison TM (2013) The cultural evolution of human communication systems in different sized populations. PLOS ONE 8(8): e71781
  2. Sulik J (2018) Cognitive mechanisms for inferring the meaning of novel signals during symbolisation. PLOS ONE 13(1): e0189540 (PMC)
  3. Hawkins RD, Sano M, Goodman ND, Fan JE (2021) Visual resemblance and communicative context constrain the emergence of graphical conventions (arXiv preprint)
  4. University of Glasgow Enlighten: Garrod S et al. (2007) Foundations of representation: where might graphical symbol systems come from? Cognitive Science 31(6)
  5. University of Glasgow Enlighten: Garrod S, Fay N et al. (2010) Can iterated learning explain the emergence of graphical symbols? Interaction Studies 11(1)

ABOUT THIS NAME

Interested in leasing doodlchat.com?

doodlchat.com is part of the DoodleSale portfolio. It isn't in an auction right now, but you can ask about leasing it.

Ask about this name