SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2109.01100 · 2021

How Suitable Are Subword Segmentation Strategies for Translating Non-Concatenative Morphology?

arXiv · PDF · Open in the Atlas

Code that ran

3 of 3 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 0 have no contract verdict; 3 failures of our instrument; 0 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/2109.01100.json · badge: /api/ran/2109.01100.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
zurichnlp/segtest canonical (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
FunctionStatusWhere it lives
get_freq_band Ran
ran; our contract draft was wrong, not the code
zurichnlp/segtest/scripts/evaluate.py
code served (permissive licence) · get_code("60db251c83c4ca5c")
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) 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 torch
import numpy as np

# The function get_freq_band takes a single parameter: freq (an int)
# It returns a string based on which frequency bucket the int falls into
# No complex objects needed - just an integer

ARGS = {
    'freq': 3
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, torch — 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_order Ran
ran; our contract draft was wrong, not the code
zurichnlp/segtest/scripts/evaluate.py
code served (permissive licence) · get_code("c917137b8314d48c")
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) 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 torch
import numpy as np
from typing import List, Dict, Tuple

COMPOUND_SURFACE_ORDER = ['sona_surface-bico', 'zogaze_surface-qepus', 'suyi_surface-saqo', 'segalo_surface-domabu', 'zarumo_surface-vazaga', 'tumoko_surface-pifeke', 'necib_surface-kixaka', 'yoyexo_surface-nifa', 'dawida_surface-nonujo', 'saxuj_surface-tedubo']

VOWEL_HARMONY_SURFACE_ORDER = ["duji_surface-['s', 'f', 'p']", "zoged_surface-['b', 'p', 'r']", "dulana_surface-['n', 'l', 'j']", "xefoqi_surface-['b', 'k', 'm']"]

INFIX_SURFACE_ORDER = ['jetah_surface-huheke', 'dezaxe_surface-siye', 'yusid_surface-huxi', 'yadey_surface-numime']

CIRCUMFIX_SURFACE_ORDER = ["['jeb', 'fet']_surface-wofi", "['Kur', 'maz']_surface-quroc", "['nuw', 'daf']_surface-seyet", "['Rül', 'bos']_surface-sudizu"]

REDUPLICATION_SURFACE_ORDER = ['partial_surface-popera', 'triple_partial_surface-metuza', 'full_surface-gija']

COMPOUND_ABSTRACT_ORDER = ['@COMPOUND_1@_abstract-wuze', '@COMPOUND_2@_abstract-pomuy', '@COMPOUND_3@_abstract-quyeso', '@COMPOUND_4@_abstract-zolo', '@COMPOUND_5@_abstract-nemine', '@COMPOUND_6@_abstract-fexot', '@COMPOUND_7@_abstract-zixu', '@COMPOUND_8@_abstract-tazif', '@COMPOUND_9@_abstract-supu', '@COMPOUND_10@_abstract-jifo']

VOWEL_HARMONY_ABSTRACT_ORDER = ['raxuja_abstract-@VOWEL_HARMONY_1@', 'gapu_abstract-@VOWEL_HARMONY_2@', 'soyut_abstract-@VOWEL_HARMONY_3@', 'zide_abstract-@VOWEL_HARMONY_4@']

INFIX_ABSTRACT_ORDER = ['@INFIX_1@_abstract-soxu', '@INFIX_2@_abstract-ceri', '@INFIX_3@_abstract-lasi', '@INFIX_4@_abstract-jigaq']

CIRCUMFIX_ABSTRACT_ORDER = ['@CIRCUMFIX_1@_abstract-fuge', '@CIRCUMFIX_2@_abstract-zixer', '@CIRCUMFIX_3@_abstract-xobex', '@CIRCUMFIX_4@_abstract-pasoz']

REDUPLICATION_ABSTRACT_ORDER = ['partial_abstract-gegec', 'triple_partial_abstract-yisu', 'full_abstract-jufo']

ORDER_LOOKUP = {'surface': {'infix': INFIX_SURFACE_ORDER, 'circumfix': CIRCUMFIX_SURFACE_ORDER, 'vowelharmony': VOWEL_HARMONY_SURFACE_ORDER, 'compound': COMPOUND_SURFACE_ORDER, 'reduplication': REDUPLICATION_SURFACE_ORDER}, 'abstract': {'infix': INFIX_ABSTRACT_ORDER, 'circumfix': CIRCUMFIX_ABSTRACT_ORDER, 'vowelharmony': VOWEL_HARMONY_ABSTRACT_ORDER, 'compound': COMPOUND_ABSTRACT_ORDER, 'reduplication': REDUPLICATION_ABSTRACT_ORDER}}

def get_order(morphological_phenomenon: str, representation_type: str) -> List[str]:
    """Returns a specific order of pattern pairs."""
    return ORDER_LOOKUP[representation_type][morphological_phenomenon]

ARGS = {
    'morphological_phenomenon': 'compound',
    'representation_type': 'surface'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, torch — 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
get_string_without_punct Ran
ran; our contract draft was wrong, not the code
zurichnlp/segtest/scripts/evaluate.py
code served (permissive licence) · get_code("5a8bb155b98c5a51")
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) 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
import string

TRANSLATOR = str.maketrans('', '', string.punctuation + '"' + '"' + "'")

def get_string_without_punct(token: str) -> str:
    """Removes all type of punctuation from a string."""
    punct_to_ignore = ["'", '-']
    for num, punct in enumerate(punct_to_ignore):
        if punct in token and (not token.endswith(punct)):
            token = re.sub('-', f'@@@{num}', token)
    token = token.translate(TRANSLATOR)
    for num, punct in enumerate(punct_to_ignore):
        if punct in token and (not token.endswith(punct)):
            token = re.sub(f'@@@{num}', '-', token)
    return token

# Test cases with various punctuation scenarios
test_token_1 = "hello-world"  # hyphen in middle
test_token_2 = "test'"  # apostrophe at end
test_token_3 = "don't"  # apostrophe in middle
test_token_4 = "simple"  # no punctuation
test_token_5 = "hello-world'"  # hyphen and apostrophe

ARGS = {
    "token": "hello-world"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 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 3 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

Data-driven subword segmentation has become the default strategy for open-vocabulary machine translation and other NLP tasks, but may not be sufficiently generic for optimal learning of non-concatenative morphology. We design a test suite to evaluate segmentation strategies on different types of morphological phenomena in a controlled, semi-synthetic setting. In our experiments, we compare how well machine translation models trained on subword- and character-level can translate these morphological phenomena. We find that learning to analyse and generate morphologically complex surface representations is still challenging, especially for non-concatenative morphological phenomena like reduplication or vowel harmony and for rare word stems. Based on our results, we recommend that novel text representation strategies be tested on a range of typologically diverse languages to minimise the risk of adopting a strategy that inadvertently disadvantages certain languages.

For agents

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

get_harvested_code_for_paper("2109.01100")
get_code_for_paper("2109.01100")
have("2109.01100")

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

curl https://syntology.ai/api/ran/2109.01100.json

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

[![Syntology run record](https://syntology.ai/api/ran/2109.01100.svg)](https://syntology.ai/paper/2109.01100)

Connect an agent — have() is free.