Dialect and regional variation studies how language differs systematically across geographic space. Isoglosses — lines on a map marking the boundary of a particular linguistic feature — bundle together to reveal dialect regions, though boundaries are typically gradient rather than sharp. In many language areas, neighboring varieties are mutually intelligible, forming a dialect continuum where the endpoints may not understand each other even though each adjacent pair can. Dialect atlases document these geographic distributions through systematic fieldwork, mapping phonological, lexical, and grammatical features across regions. The distinction between "language" and "dialect" is ultimately socio-political rather than purely linguistic, famously captured in the quip that a language is a dialect with an army and a navy.
Study a dialect atlas (like the Atlas of North American English or the Survey of English Dialects) and trace specific isoglosses — the cot-caught merger, rhoticity, or the Northern Cities Vowel Shift. Listen to recordings from different regions and try to identify the phonological features that distinguish them. Map a feature you notice varying in your own region to build intuition for fieldwork methodology.
From your study of sociolinguistics, you know that language variation is not random noise but patterned along social dimensions — class, age, ethnicity, gender. Dialect geography extends this into the spatial dimension: language also varies systematically across physical space, and the patterns of that variation can be mapped. The central tool for mapping linguistic geography is the isogloss — a line drawn on a map marking the approximate boundary between two regions where speakers use different forms of the same feature. You might draw an isogloss separating the region where "cot" and "caught" are pronounced identically from the region where they're distinct, or separating the area where "pop" is used for soft drinks from the area where "soda" dominates.
The critical insight is that isoglosses for different features rarely coincide. The boundary for one vowel merger runs through one set of counties; the boundary for a different lexical item runs through a different set. When many isoglosses cluster together in roughly the same area, that cluster defines a dialect region — a zone where multiple features distinguish the speech on one side from the speech on the other. The American Midwest, the American South, New England, and other recognized dialect areas are defined not by single features but by bundles of isoglosses that happen to concentrate in the same transition zone. Where isoglosses are spread out, transitions are gradual; where they cluster sharply, dialect boundaries feel more real.
This leads to the concept of the dialect continuum: in many large language areas, every adjacent pair of communities can understand each other reasonably well, but the varieties at opposite ends of the continuum may be mutually unintelligible. Medieval European peasants could travel from one village to the next without hitting a communication wall — but a speaker from northern Germany and a speaker from southern Italy were effectively speaking different languages, even though they were connected through an unbroken chain of intelligible steps. The Arabic-speaking world is a modern example: Egyptian Arabic and Moroccan Arabic speakers may struggle to understand each other, yet the dialects between them form a continuum of gradual change.
This fact dissolves the question "is X a language or a dialect?" The famous quip — a language is a dialect with an army and a navy — captures the real answer: the distinction is political, not linguistic. Norwegian and Swedish are mutually intelligible and might be called dialects of a single Scandinavian speech variety, but political history made them "separate languages." Serbian, Croatian, and Bosnian are largely mutually intelligible but are now officially separate languages for national identity reasons. Mandarin and Cantonese are officially "dialects of Chinese" but are not mutually intelligible. The linguistic facts don't determine the label; social and political facts do. For linguists, the neutral term variety sidesteps the loaded language/dialect dichotomy and lets the structural analysis proceed without prejudging the political question.