Hebrew OT Predicate-Argument Structure (Semantic Role Labeling)¶

Analysis of the MACULA Hebrew WLC frame column, which encodes PropBank-style semantic role structure for ~68,207 Hebrew verb tokens (~14.6% of all Hebrew OT tokens).

Argument roles:

• A0 — proto-agent: the subject / initiator of the action
• A1 — proto-patient: the object / affected entity

Questions this notebook answers:

• What does אֱלֹהִים (God) do as agent across the OT?
• What is created/destroyed/given as patient across the OT?
• Who are the typical agents of theological verbs like בָּרָא (create), נָתַן (give), שָׁפַט (judge)?
• What does YHWH act on most often?
• What are the most common agent–verb–patient triples in each book?

Data source: MACULA Hebrew WLC (macula-hebrew/ submodule), semantic roles from Clear Bible / Groves Center.

In [ ]:

import sys
sys.path.insert(0, '../../../src')

from bible_grammar import (
    ot_frame_data, ot_agent_verbs, ot_patient_verbs,
    ot_verb_agents, ot_verb_patients, ot_frame_pairs,
    print_ot_frame_overview, print_ot_agent_verbs, print_ot_patient_verbs,
    print_ot_verb_agents, print_ot_verb_patients, print_ot_frame_pairs,
    ot_agent_verbs_chart, ot_patient_verbs_chart,
)
import pandas as pd

import sys sys.path.insert(0, '../../../src') from bible_grammar import ( ot_frame_data, ot_agent_verbs, ot_patient_verbs, ot_verb_agents, ot_verb_patients, ot_frame_pairs, print_ot_frame_overview, print_ot_agent_verbs, print_ot_patient_verbs, print_ot_verb_agents, print_ot_verb_patients, print_ot_frame_pairs, ot_agent_verbs_chart, ot_patient_verbs_chart, ) import pandas as pd

1. Overview¶

In [ ]:

print_ot_frame_overview()

print_ot_frame_overview()

In [ ]:

# Sample frame data
df = ot_frame_data(book='Gen')
print(f"Genesis frame-annotated verbs: {len(df)}")
df[['text', 'lemma', 'gloss', 'a0_lemma', 'a0_gloss', 'a1_lemma', 'a1_gloss',
    'chapter', 'verse']].head(15)

# Sample frame data df = ot_frame_data(book='Gen') print(f"Genesis frame-annotated verbs: {len(df)}") df[['text', 'lemma', 'gloss', 'a0_lemma', 'a0_gloss', 'a1_lemma', 'a1_gloss', 'chapter', 'verse']].head(15)

2. What Does God Do? — Elohim and YHWH as Agent¶

The A0 (agent) slot answers: who performs the action? Looking at אֱלֹהִים and יהוה as A0 reveals the OT's portrait of divine activity.

In [ ]:

# What does אֱלֹהִים (God / Elohim) do as agent?
print_ot_agent_verbs('אֱלֹהִים', top_n=20)

# What does אֱלֹהִים (God / Elohim) do as agent? print_ot_agent_verbs('אֱלֹהִים', top_n=20)

In [ ]:

# What does יהוה (YHWH) do as agent?
print_ot_agent_verbs('יהוה', top_n=20)

# What does יהוה (YHWH) do as agent? print_ot_agent_verbs('יהוה', top_n=20)

In [ ]:

ot_agent_verbs_chart('יהוה', top_n=15)

ot_agent_verbs_chart('יהוה', top_n=15)

3. What Acts on Israel / the People? — Patient Analysis¶

The A1 (patient) slot answers: what receives the action?

In [ ]:

# What verbs take 'Israel' as patient?
print_ot_patient_verbs('יִשְׂרָאֵל', top_n=15)

# What verbs take 'Israel' as patient? print_ot_patient_verbs('יִשְׂרָאֵל', top_n=15)

In [ ]:

# What verbs take 'the people' (עַם) as patient?
print_ot_patient_verbs('עַם', top_n=15)

# What verbs take 'the people' (עַם) as patient? print_ot_patient_verbs('עַם', top_n=15)

4. Theological Verb Profiles¶

For key theological verbs, we can ask: who does this action and what is acted upon?

In [ ]:

# בָּרָא (create) — who creates? what gets created?
print_ot_verb_agents('בָּרָא', top_n=10)
print_ot_verb_patients('בָּרָא', top_n=10)

# בָּרָא (create) — who creates? what gets created? print_ot_verb_agents('בָּרָא', top_n=10) print_ot_verb_patients('בָּרָא', top_n=10)

In [ ]:

# נָתַן (give) — who gives? what gets given?
print_ot_verb_agents('נָתַן', top_n=10)
print_ot_verb_patients('נָתַן', top_n=10)

# נָתַן (give) — who gives? what gets given? print_ot_verb_agents('נָתַן', top_n=10) print_ot_verb_patients('נָתַן', top_n=10)

In [ ]:

# שָׁפַט (judge) — who judges? what/who gets judged?
print_ot_verb_agents('שָׁפַט', top_n=10)
print_ot_verb_patients('שָׁפַט', top_n=10)

# שָׁפַט (judge) — who judges? what/who gets judged? print_ot_verb_agents('שָׁפַט', top_n=10) print_ot_verb_patients('שָׁפַט', top_n=10)

In [ ]:

# בָּרַךְ (bless) — who blesses? who gets blessed?
print_ot_verb_agents('בָּרַךְ', top_n=10)
print_ot_verb_patients('בָּרַךְ', top_n=10)

# בָּרַךְ (bless) — who blesses? who gets blessed? print_ot_verb_agents('בָּרַךְ', top_n=10) print_ot_verb_patients('בָּרַךְ', top_n=10)

In [ ]:

# שָׁמַע (hear/obey) — who hears? what is heard?
print_ot_verb_agents('שָׁמַע', top_n=10)
print_ot_verb_patients('שָׁמַע', top_n=10)

# שָׁמַע (hear/obey) — who hears? what is heard? print_ot_verb_agents('שָׁמַע', top_n=10) print_ot_verb_patients('שָׁמַע', top_n=10)

5. Most Common Agent–Verb–Patient Triples¶

In [ ]:

# Top triples across the entire OT
print_ot_frame_pairs(top_n=25)

# Top triples across the entire OT print_ot_frame_pairs(top_n=25)

In [ ]:

# Top triples in Genesis
print_ot_frame_pairs(book='Gen', top_n=15)

# Top triples in Genesis print_ot_frame_pairs(book='Gen', top_n=15)

In [ ]:

# Top triples in Deuteronomy (Torah command context)
print_ot_frame_pairs(book='Deu', top_n=15)

# Top triples in Deuteronomy (Torah command context) print_ot_frame_pairs(book='Deu', top_n=15)

In [ ]:

# Top triples in Isaiah (prophetic context)
print_ot_frame_pairs(book='Isa', top_n=15)

# Top triples in Isaiah (prophetic context) print_ot_frame_pairs(book='Isa', top_n=15)

6. Ad-hoc Queries¶

In [ ]:

# What does Moses (מֹשֶׁה) do as agent?
print_ot_agent_verbs('מֹשֶׁה', top_n=15)

# What does Moses (מֹשֶׁה) do as agent? print_ot_agent_verbs('מֹשֶׁה', top_n=15)

In [ ]:

# What is given (נָתַן) vs created (בָּרָא) — compare patients
given = ot_verb_patients('נָתַן', top_n=10)
created = ot_verb_patients('בָּרָא', top_n=10)

print("נָתַן patients:")
print(given[['a1_lemma', 'a1_gloss', 'count']].to_string(index=False))
print()
print("בָּרָא patients:")
print(created[['a1_lemma', 'a1_gloss', 'count']].to_string(index=False))

# What is given (נָתַן) vs created (בָּרָא) — compare patients given = ot_verb_patients('נָתַן', top_n=10) created = ot_verb_patients('בָּרָא', top_n=10) print("נָתַן patients:") print(given[['a1_lemma', 'a1_gloss', 'count']].to_string(index=False)) print() print("בָּרָא patients:") print(created[['a1_lemma', 'a1_gloss', 'count']].to_string(index=False))

In [ ]:

# Raw frame data for a specific passage (Genesis 1)
gen1 = ot_frame_data(book='Gen')
gen1_ch1 = gen1[gen1['chapter'] == 1]
gen1_ch1[['text', 'lemma', 'gloss', 'a0_lemma', 'a0_gloss', 'a1_lemma', 'a1_gloss', 'verse']]

# Raw frame data for a specific passage (Genesis 1) gen1 = ot_frame_data(book='Gen') gen1_ch1 = gen1[gen1['chapter'] == 1] gen1_ch1[['text', 'lemma', 'gloss', 'a0_lemma', 'a0_gloss', 'a1_lemma', 'a1_gloss', 'verse']]