Special issue of Pawari
Shodh Patrika
Study U-1
Bayesian Time-Calibrated Phylogeny of
Bhoyari (Pawari): BEAST2, PPC and ABC Evidence for Rajasthani Origin
Bayesian Phylogenetic
Network, Posterior Predictive Checking,
and Approximate Bayesian Computation — A Complete Methodological
Framework
ISBN: 978-620-9-62547-3
Author
Shivani Barange Pawar
Researcher
– Bhoyari / Pawari Language Researcher, Folk Culture and Satpuda Regional
Studies
Co-Editor,
Writer & Contributor – Maa Tapti Shodh Sansthan, Multai, Betul & Pawari
Shodh Patrika
Rajesh Barange Pawar
Independent
Researcher | Bhoyari / Pawari Language Researcher | Pawar Bhoyar Community
History | Rajasthan Malwa Satpuda Regional Studies
Founder
& Director – Maa Tapti Shodh Sansthan, Multai, Betul | Chief Editor –
Pawari Shodh Patrika
Pranay Chopde
Independent
Researcher | Pawar Community History | Rajasthan Malwa Satpuda Regional Studies
Co-Founder–
Maa Tapti Shodh Sansthan, Multai, Betul | Editor – Pawari Shodh Patrika
Published by
Maa Tapti Shodh Sansthan,
Multai, Betul, Madhya Pradesh
ABSTRACT
This study presents a Bayesian time-calibrated phylogenetic analysis of
Satpuda Bhoyari's divergence within the Western Indo-Aryan macro-family, using
three complementary computational historical linguistics frameworks: (1) BEAST2
Bayesian phylogenetic inference with a Relaxed Lognormal Clock and Binary
Covarion substitution model; (2) Bayesian Posterior Predictive Checking (PPC)
for model adequacy assessment; and (3) Approximate Bayesian Computation (ABC)
for independent divergence time validation. The input data is a 200-item binary
cognate matrix constructed from the authenticated Satpuda Bhoyari Lokgeet
corpus and lexical dataset across six taxa (Bhoyari, Marwari, Mewadi, Rangdi,
Malvi, Bundeli). Historical calibration priors are anchored to documented
Western Indo-Aryan attestation dates.
BEAST2 analysis estimates Bhoyari divergence from Core Rajasthani at a
posterior mean of 700 years before present (≈ 1325 CE), with a 95% Highest
Posterior Density (HPD) interval of 575–825 BP (≈ 1200–1450 CE). Posterior node
probability at the Proto-Bhoyari node is 0.92 — strong phylogenetic support.
The relaxed clock identifies a three-phase trajectory: Proto-Bhoyari founding
(~1300–1400 CE), Malwa contact phase (~1500–1700 CE), and Satpuda stabilization
(~1700 CE onward). The Bayesian Skyline model shows a population proxy dip at
founding, followed by gradual expansion during the Malwa phase and
stabilization plateau post-1700 CE. PPC confirms adequate model fit: all three
summary statistics (mean pairwise Hamming distance, homoplasy index, normalized
tree height) fall within the 95% simulated range. ABC independently confirms
the divergence estimate: posterior mean ≈ 680 BP, with Model A (~1300 CE split)
receiving posterior probability 0.71 vs Model B (~1500 CE split, P=0.19) and
Model C (~1700 CE split, P=0.10). SplitsTree5 phylogenetic network analysis
reveals moderate reticulation (δ = 0.21) confirming Malwa contact influence
without altering primary Rajasthani inheritance. Five independent computational
methods converge on a single conclusion: Satpuda Bhoyari diverged from Core
Rajasthani ~1300–1400 CE and was never derived from Malvi.
Satpuda
Bhoyari (Pawari) exhibits lexical and phonological similarities with Malvi as a
result of an extended Malwa contact phase lasting approximately 250–300 years
during historical migration and settlement processes. However, these
similarities are primarily superficial and arise from areal convergence rather
than genetic affiliation.
Subsequent
linguistic surveys and regional classifications, relying largely on such
surface features, placed Bhoyari within the Malvi subgroup. Nevertheless, a
deeper structural analysis—particularly of its phonology, morphology, and core
grammatical system—reveals a significantly closer alignment with the Core
Rajasthani linguistic group.
Therefore,
Bhoyari is more appropriately understood as a southern extension of the
Rajasthani dialect continuum, and its classification under Malvi represents a
case of contact-induced misclassification rather than true linguistic
derivation.
Keywords: BEAST2,
Bayesian phylogenetics, time-calibrated tree, relaxed lognormal clock, Satpuda
Bhoyari, Western Indo-Aryan, posterior predictive checking, approximate
Bayesian computation, phylogenetic network, SplitsTree5, divergence time,
historical linguistics, migration-chain model, Malwa contact, cognate matrix
Bhoyari/Pawari (Satpuda Zone) is a
sub-dialect of Malvi (as per linguistic researcher), belonging to the Western
Indo-Aryan branch of the Indo-European language family. It is spoken
exclusively by the Pawar / Bhoyar Pawar /Kshatriya Bhoyar Pawar community in
the Satpuda region — specifically in Betul district, Chhindwara district, the
Pandhurna region, and Wardha, in the states of Madhya Pradesh and Maharashtra.
The dialect exhibits a rare sociolinguistic property: it is simultaneously
geographically bounded and community-exclusive. No speakers of other castes or
communities use Bhoyari/Pawari (Satpuda Zone) as a primary communicative code.
The Betul–Chhindwara–Pandhurna triangle is
considered the prestige core of the dialect. The variety spoken in this
triangle is regarded by community scholars and external observers alike as the
most authentic and phonologically conservative form of Bhoyari/Pawari. The
Wardha variety shows greater Marathi contact influence, while the
Betul–Chhindwara core retains stronger Western Indo-Aryan structural features.
FIGURES
— BAYESIAN PHYLOGENETIC ANALYSIS
The eight
figures below provide complete visual coverage of the Study U-1 analytical
results, from posterior density estimation through phylogenetic network
visualization, skyline modeling, and ABC validation.
|
Figure 1. Bayesian
Posterior Density — Bhoyari Divergence Time
BEAST2 posterior density (years before present). Green shading
= core credible interval (650–750 BP). Blue shading = 95% HPD (575–825 BP ≈
1200–1450 CE). Gold dashed = posterior mean 700 BP (~1325 CE). Upper axis
shows corresponding CE dates. Unimodal distribution with slight right skew —
typical of time-calibrated analyses. ESS > 250 confirms adequate sampling.
|
|
Figure 2.
Time-Calibrated Bayesian Phylogenetic Tree
Maximum Clade Credibility tree from BEAST2 with posterior
probabilities. X-axis = years before present (right = present). Proto-Bhoyari
node (gold diamond) at ~1325 CE, PP=0.92. Dashed orange arrows = Malwa
contact edges. Bundeli (red dotted) on separate Central IA branch. Three
migration phases labeled: founding (~1325 CE), Malwa phase (~1500–1700 CE),
Satpuda (~1700+).
|
|
Figure 3. Bayesian
Skyline Population Proxy — Migration Phase Modeling
Normalized effective speaker population proxy (Ne) across
Bhoyari history. Blue band = 95% credible interval. Three annotated phases:
dip at Proto-Bhoyari founding (~1325 CE, small group); gradual expansion
during Malwa contact (~1500–1700 CE); stable plateau post-1700 CE (Satpuda
stabilization). Pattern consistent with founder effect → growth →
stabilization.
|
|
Figure 4. Bayesian
Posterior Predictive Checks (PPC) — Three Summary Statistics
Simulated distributions (blue histograms, n=1000) vs observed
statistics (red lines). Left: Mean pairwise Hamming distance (obs=0.28, sim
mean=0.27, range 0.22–0.31). Centre: Homoplasy index (obs=0.14, sim
mean=0.13). Right: Normalized tree height (obs=0.72). All three observed
values fall within 95% simulated range — model fit confirmed (✓).
|
|
Figure 5. Approximate
Bayesian Computation (ABC) — Divergence Model Comparison
Left: ABC posterior density (blue, top 1% of 50,000
simulations) vs uniform prior (gray). ABC mean ≈ 680 BP (gold dashed),
consistent with BEAST2 estimate (red dotted, 700 BP). Right: Model comparison
— Model A (~1300 CE split) receives posterior probability 0.71; Models B and
C receive 0.19 and 0.10. Two independent Bayesian frameworks agree.
|
|
Figure 6. Bayesian
Phylogenetic Network — SplitsTree5 NeighborNet
Primary tree edges (solid colored lines) represent vertical
inheritance. Dashed curved lines = contact/reticulation edges. Orange curve:
Bhoyari ↔ Malvi (Malwa phase contact, δ=0.21). Purple curve: Bhoyari ↔ Rangdi
(weak contact, δ=0.16). Marwari and Mewadi show low δ-scores (0.07–0.08),
confirming tree-like inheritance. No reticulation edge with Bundeli —
confirms separate Central IA lineage.
|
|
Figure 7. Lexical
Retention Decay Model — L(t) = e^(−λt)
Left: Decay curves for all six varieties, calibrated λ values
shown. Bhoyari (gold, λ=0.0011) retains ~46% of conservative core at 700 BP —
matches observed corpus similarity of 40–55%. Right: Bhoyari λ posterior
distribution. Prior (gray) = Normal(0.0011, 0.0003). Posterior (blue)
narrowed by likelihood — ML estimate λ=0.0011 with 95% CI: 0.00095–0.00125
per year.
|
|
Figure 8. Combined
Model Validation Radar — Five Computational Methods
Five-axis radar comparing performance of all analytical
frameworks across: posterior probability, temporal accuracy, PPC model fit,
ABC confirmation, and phylogenetic support. All four methods (green, teal,
blue, purple) show consistently high scores across all five dimensions —
convergent validation of the ~1300 CE divergence estimate.
|
The present
linguistic and computational analysis not only establishes the phylogenetic
position of Satpuda Bhoyari (Pawari), but also provides strong indirect
historical evidence regarding the migration trajectory of the Bhoyar Pawar
community.
The convergence
of five independent Bayesian methods—BEAST2, Bayesian Skyline, Posterior
Predictive Checking, Approximate Bayesian Computation, and SplitsTree5—confirms
that Bhoyari diverged from the Core Rajasthani linguistic group around
1300–1400 CE. This linguistic divergence is not merely a structural phenomenon;
it reflects a real historical separation of a speech community from the
Rajasthani region.
Crucially, the
presence of a prolonged Malwa contact phase (approximately 250–300 years),
supported by measurable contact parameters (δ ≈ 0.21), lexical borrowing
patterns, and demographic expansion signals, indicates that this migrating
population did not move directly to the Satpuda region. Instead, it passed
through and remained in the Malwa region for several generations, allowing
significant but surface-level linguistic interaction with Malvi.
This
sequence—Rajasthani origin → Malwa contact → Satpuda stabilization—is not an
assumption but a model consistently supported by multiple independent datasets
and analytical frameworks.
Therefore, the
linguistic evidence aligns closely with and strongly supports the historical
migration narrative of the Bhoyar Pawar community:
Rajasthani
Origin (~1300 CE):
The divergence
of Bhoyari from Core Rajasthani indicates that the ancestral Bhoyar Pawar
population was originally part of the Rajasthani linguistic and cultural
sphere.
Malwa Phase
(~1400–1700 CE):
The
statistically detectable Malwa contact phase demonstrates a prolonged
settlement or interaction period in the Malwa region, during which linguistic
convergence with Malvi occurred.
Satpuda
Settlement (~1700 CE onwards):
The
stabilization phase marks the final establishment of the community in the
Satpuda region (Betul–Chhindwara zone), where Bhoyari developed into its
present form.
Importantly,
this linguistic reconstruction provides independent scientific validation of
the community’s migration history—separate from oral traditions, genealogies,
or historical texts. In other words, the language itself preserves the
migration pathway.
Key Inference
The Bhoyari
language functions as a linguistic record of migration: it proves that the
Bhoyar Pawar community originated in Rajasthan, underwent a significant Malwa
phase, and later settled in the Satpuda region.
Thus, this study
goes beyond classification. It establishes that:
Bhoyari is
structurally Rajasthani — confirming origin
Malvi
influence is contact-based — confirming Malwa phase
Final
linguistic stabilization in Satpuda — confirming settlement
Final
Statement
Language here is
not just a medium of communication—it is historical evidence.
And in the case
of Bhoyari, it clearly encodes the Rajasthan → Malwa → Satpuda migration
pathway of the Bhoyar Pawar community.
|
Five methods | One conclusion: Bhoyari
← Core Rajasthani ~1300 CE | NOT from Malvi | But Classified under Malvi | Satpuda
stabilization ~1700 CE | PP = 0.92
|
