Identifying Dialect Regions
Specific features vs. overall measures
using the Romanian Online Dialect Atlas
and Multidimensional Scaling
• Background
– Romanian
• Dialects: England, Finland and Romania
• What makes a dialect?
Romanian and Romance
RODA: Romanian Online Dialect Atlas
or temporarily at under research
RODA: Technology
multi-dimensional scaling
England (CLAE)
Geography ~ Linguistics
• In both the English and the Finnish cases,
“geography” closely maps to “dialect distance”:
– Geographically-distant regions are dialect-distant
– Geographically-compact areas are dialect-compact
– Overall arrangement of areas is (mostly) the same in
both cases
No Surprise: geography is a factor in dialectology
and the exceptions have explanations
North-West Romania
• In North-west Romania, this is not the case.
– There are two areas in the south that are suitably
– The rest overlay one another in ways that do not
match geography
– Areas that we expect to be separate are not, e.g.:
• Oaş divides into two areas
• Both are mixed in with other dialect areas.
North West Romania
North West Romania (part)
Oaş (non-palatalized dentals)
Oaş (1)
Oaş (2)
Oaş (3)
What makes a dialect area?
• Selected features
show Oaş as a
separate dialect area
• MDS on “all data”
shows Oaş as two
areas, intermixed
with the
surrounding areas
What is a Dialect?
Different features- Different maps
• Technically, we can
overlay several
patterns on one map
• So, where are the
dialect regions?
(I) Distinct dialect regions
• Each dialect region is
defined by a unique
set of features
• Even though some of
the features are shared
from region to region
(II) Overlapping dialect regions
• Or, do we keep only 3 overlapping dialect
Multiple patterns
The question arises over and over because:
• Every feature has its own pattern
• Many patterns coincide in some locations
and not in others
• Features that are quantitative have
potentially infinite possibilities
How do we put these patterns together
Proposal: Dialect Structure
Hierarchy of dialect patterns
Dialect Structure
We propose to speak of the dialect structure
• Multiple levels
• Dialect region exists at one (or more) level(s)
• At different levels, there are different patterns
of dialect regions
It is the total “dialect structure” that is
essential, rather than any one level or view
of it.
Quantitative Measures: Raised, word-final /e/ vs schwa
Data from:
•407 maps
•Field 1
Raised /e/
Raised schwa
Raised schwa is
also wide-spread
but does not
always coincide
with raised /e/
(cf. 158, 159)
Quantitative Measures
• Is there vowel raising?
(Answer can cover many
degrees of raising)
• How many occurrences
of syllabic /u/ vs. nonsyllabic /u/ (Answers
range from 0 to 400+)
Quantity in Dialect Structure
At any level of dialect structure, the
distinction between regions can be variable
because of the quantitative nature of what is
being measured and where we choose to set
the threshold.
This threshold is part of the “dialect structure”
Dialect Structure
• Hierarchy of levels
– Ground level, intermediate levels, top level
• Thresholds for variable measures
• Views of the whole structure
– View = dialect map at a given level and threshold
MDS can provide one top-level view, where
every measured features has equal weight
North West Romania
Reality of the Dialect Structure
Does the multi-level, quantitatively variable
structure correspond to something real?
Or, are the patterns just the artifacts of waves
of change moving through the speech
community, overlaid by new patterns
through time
Reality: England, Finland
Distinctions seem to be:
• Clear cut
• Persistent
• Explained by geography (mostly)
So perhaps the dialect structure is “real”
Reality: Romania
• In Daco-Romanian, most distinguishing
features are phonetic or lexical
• Fewer morphological or syntactic
• If the dialect areas are “real”, will they
eventually also show more differences
including differences in morphology and
• RODA is up and working
– We have used it to investigate interesting
problems in Romance and Romanian
• Our investigations have forced us to ask
“what is a dialect?”
• Our proposal: A “dialect structure” with
multiple levels and quantitative thresholds
– Open Question: is it real?
RODA: Romanian Online Dialect Atlas
or temporarily at under research

Identifying Dialect Regions