SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2010.02069 · EMNLP · 2020

The Grammar of Emergent Languages

Dieuwke Hupkes, Oskar Van Der Wal, Elia Bruni, Silvan De Boer

arXiv · PDF · Open in the Atlas

Code that ran

5 of 7 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 5 have no contract verdict; 0 failures of our instrument; 2 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2010.02069.json · badge: /api/ran/2010.02069.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
i-machine-think/emergent_grammar_induction canonical (classifier label)
classifier basis: census
5 ran · 2 no recorded run
FunctionStatusWhere it lives
find_message Ran
ran on a synthesized input
i-machine-think/emergent_grammar_induction/utils/convert2constituents.py
code served (permissive licence) · get_code("3b08575d0a89e152")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned str. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import re

def find_message(tree, shapes=False):
    """ Find the message in plain text from the tree """
    if not shapes:
        return ' '.join(re.findall('\\d+', tree)) + ' .'
    else:
        return ' '.join(re.findall('[^\\s\\(\\)]+', tree)) + ' .'

# Setup code to construct valid arguments
tree_example_1 = "node(1, 2, 3) leaf(4, 5)"
tree_example_2 = "(alpha beta) (gamma delta) 123"

ARGS = {
    "tree": tree_example_1,
    "shapes": False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_terminals Ran
ran on a synthesized input
i-machine-think/emergent_grammar_induction/utils/analysis.py
code served (permissive licence) · get_code("8e46c475b0ce9d8a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import nltk
from nltk import CFG, Production, Nonterminal

# Create a simple PCFG with some productions
cfg_string = """
    S -> NP VP
    NP -> Det N
    VP -> V NP
    Det -> 'the' | 'a'
    N -> 'dog' | 'cat'
    V -> 'chased' | 'saw'
"""

pcfg = CFG.fromstring(cfg_string)

ARGS = {
    'pcfg': pcfg
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
nltk — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
remove_redundant_brackets Ran
ran on a synthesized input
i-machine-think/emergent_grammar_induction/utils/convert2constituents.py
code served (permissive licence) · get_code("2f7607436091dd6b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned str. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import re

def remove_redundant_brackets(span):
    """ Removes the bracketing in the tree for 0-1 ... (n-1)-n once.
    Because extra brackets were added to read the tree properly in NLTK,
    these brackets need to be removed again """
    result = [e for e in re.split('[^0-9]', span) if e != '']
    max_int = max(map(int, result))
    span = span.split()
    newspan = []
    ranges = [f'{i}-{i + 1}' for i in range(0, max_int)]
    for j in span:
        if j in ranges:
            ranges.remove(j)
            continue
        else:
            newspan.append(j)
    return ' '.join(newspan)


# Test case 1: Simple span with redundant brackets
span_input_1 = "( 0-1 1-2 ( 2-3 ) )"

# Test case 2: Span with mixed content
span_input_2 = "( NP 0-1 ( VP 1-2 2-3 ) 3-4 )"

ARGS = {
    "span": span_input_1
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
replace_terminals_with_indices Ran
ran on a synthesized input
i-machine-think/emergent_grammar_induction/utils/convert2constituents.py
code served (permissive licence) · get_code("7ad469b797c1e955")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ParentedTree. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

from nltk import Tree, ParentedTree

# Create a simple tree string with terminals (words) that will be replaced with indices
# The format is: (NP (DET the) (N cat))
treestring = "(S (NP (DET the) (N cat)) (VP (V sat)))"

ARGS = {
    "treestring": treestring
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
nltk — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ParentedTree — a ParentedTree output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
sample_message Ran
ran on a synthesized input
i-machine-think/emergent_grammar_induction/utils/analysis.py
code served (permissive licence) · get_code("219bd6d752ce6d9a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import random

# Set seed for reproducibility
random.seed(0)

# Define the arguments
L = 5  # Length of message (number of symbols to sample)
vocabulary = ['a', 'b', 'c', 'd', 'e']  # A small vocabulary of symbols

ARGS = {
    'L': L,
    'vocabulary': vocabulary,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
overgeneration_coverage Not yet run
unverified (harvested, no recorded run)
i-machine-think/emergent_grammar_induction/utils/analysis.py
code served (permissive licence) · get_code("d5c95aebe631c210")
How it ran

no invocation recorded.

parse_induced_grammar Not yet run
unverified (harvested, no recorded run)
i-machine-think/emergent_grammar_induction/utils/bmm_labels2grammar.py
code served (permissive licence) · get_code("0e8ee009897e7801")
How it ran

no invocation recorded.

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 7 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

In this paper, we consider the syntactic properties of languages emerged in referential games, using unsupervised grammar induction (UGI) techniques originally designed to analyse natural language. We show that the considered UGI techniques are appropriate to analyse emergent languages and we then study if the languages that emerge in a typical referential game setup exhibit syntactic structure, and to what extent this depends on the maximum message length and number of symbols that the agents are allowed to use. Our experiments demonstrate that a certain message length and vocabulary size are required for structure to emerge, but they also illustrate that more sophisticated game scenarios are required to obtain syntactic properties more akin to those observed in human language. We argue that UGI techniques should be part of the standard toolkit for analysing emergent languages and release a comprehensive library to facilitate such analysis for future researchers.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2010.02069")
get_code_for_paper("2010.02069")
have("2010.02069")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2010.02069.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2010.02069.svg)](https://syntology.ai/paper/2010.02069)

Connect an agent — have() is free.