Older Southern American English

After the Civil War, the growth of timber, coal, railroad, steel, textile, and tobacco mill industries throughout the South, along with the whole country's resulting migration changes, likely contributed to the expansion of a more unified Southern accent (now associated with the 20th century), which gradually ousted 19th-century Southern accents.[7] Before World War II, the demographic tendency of the South was out-migration, but after the war, a counter-tendency emerged. Now, a high in-migration of Northerners, especially toward urban areas of the South, may have been another motivation for the abandonment of older Southern accent features. Finally, the Civil Rights Movement seems to have led white and black Southerners alike to resist accent features associated with the other racial group and even develop newly distinguishing features, which may explain the sudden adoption of rhoticity among most white Southerners (along with continued non-rhoticity among black Americans) since the middle of the 20th century onwards.[7]

The phonologies of early Southern English in the United States were diverse. The following pronunciation features were very generally characteristic of the older Southern region as a whole:

• Lack of Yod-dropping: Pairs like do and due, or toon and tune, were often distinct in these dialects because words like due, lute, new, etc. historically contained a diphthong similar to /ju/ (like the you sound in cute or puny).[9] (as England's RP standard pronunciation still does), but Labov et al. report that the only Southern speakers who make a distinction today use a diphthong /ɪu/ in such words.[10] They further report that speakers with the distinction are found primarily in North Carolina and northwest South Carolina, and in a corridor extending from Jackson to Tallahassee. For most of the South, this feature began disappearing after World War II.[11]
  • Yod-coalescence: Words like dew were pronounced as "Jew", and Tuesday as "choose day."
• Wine–whine distinction: distinction between "w" and "wh" in words like "wine" and "whine", "witch and "which", etc.
• Horse–hoarse distinction: distinction between pairs of words like "horse" and "hoarse", "for" and "four", etc.
• Rhoticity and non-rhoticity: The pronunciation of the r sound only before or between vowels, but not after vowels, is known as non-rhoticity and was historically associated with the major plantation regions of the South: specifically, the entire Piedmont and most of the South's Atlantic Coast in a band going west towards the Mississippi River, as well as all of the Mississippi Embayment and some of the western Gulf Coastal Plain. This was presumably influenced by the non-rhotic East Anglia and London England pronunciation. Additionally, some older Southern dialects were even "variably non-rhotic in intra-word intervocalic contexts, as in carry [kʰæi]."[8] Rhotic accents of the older Southern dialects, which fully pronounce all historical r sounds, were somewhat rarer and primarily spoken in Appalachia, the eastern Gulf Coastal Plain, and the areas west of the Mississippi Embayment.[12]
• Palatalization of /k/ and /g/ before /ɑr/: Especially in the older South along the Atlantic Coast, the consonants /k/ (as in key or coo) and /ɡ/ (as in guy or go), when before the sound /ɑːr/ (as in car or barn), were often pronounced with the tongue fronted towards the hard palate. Thus, for example, garden in older Southern was something like "gyah(r)den" [ˈgjɑ(ɹ)dən] and "cart" like "kyah(r)t" [cʰjɑ(ɹ)t]. This pronunciation feature was in decline by the late 1800s.[11]
• Lack or near-lack of /aɪ/ glide weakening: The gliding vowel in words like prize (but less commonly in price or other situations of this vowel appearing before a voiceless consonant) commonly has a "weakened" glide today in the South; however, this only became a documented feature since the last quarter of the 1800s and was otherwise absent or inconsistent in earlier Southern dialects. Today, the lack of glide weakening persists in the High Tider and updated Lowcountry accents. Full weakening has become a defining feature only of the modern Southern dialects, particularly the most advanced sub-varieties.[13]
• Mary–marry–merry distinction: Unlike most of the U.S. and modern Southern, older Southern did not merge the following three vowels before /r/: [e~eə] (as in Mary), [æ] (as in marry), and [ɛ] (as in merry). Although the three are now merging or merged in modern Southern English, the "marry" class of words remains the least likely among modern Southerners to merge with the other two.[14]
• Clear /l/ between front vowels: Unlike modern Southern and General American English's universally "dark" /l/ sound (often represented as [ɫ]), older Southern pronunciation had a "clear" (i.e. non-velarized) /l/ sound whenever /l/ appears between front vowels, as in the words silly, mealy, Nellie, etc.[11]
• Was, what and of pronounced with [ɑ]: The stressed word what, for example, rhymed with cot (not with cut, as it does elsewhere in the U.S.).[15]
• No happy-tensing: The final vowel of words like happy, silly, monkey, parties, etc. were not tensed as they are in newer Southern and other U.S. dialects, meaning that this vowel sounded more like the [ɪ] of fit than the [i] of feet.
• /oʊ/, as in goat, toe, robe, etc., kept a back starting place (unlike most Southern since World War II, but like most Northern U.S. dialects today); this became an opener [ɔu~ɒu] in the early 1900s.[15] The modern fronted form of the Atlantic South started as far back as the 1800s in northeastern North Carolina, in the form [ɜy], but only spread slowly, until accelerating after World War II.[16]
• /ʃr/ pronounced as [sɹ] (e.g. causing shrimp, shrub, etc. to sound like srimp, srub, etc.); this feature was reported earliest in Virginia.[17]

The area in dark purple approximates the Plantation Southern dialect region, excluding the Lowcountry (the Atlantic coast of South Carolina and Georgia).

Older speech of the Plantation South included those features above, plus:

• Non-rhoticity: R-dropping historically occurred in the greater central sections of Georgia, Alabama, and Mississippi, and in coastal Texas and some other coastal communities of the Gulf states. Rhoticity (or r-fulness) was more likely in the southernmost sections of Georgia, Alabama, and Mississippi, as well as in northern Florida, western Louisiana, and eastern Texas.[18]
• Trap–bath split: Words like bath, dance, and ask, used a different vowel ([æ̈ɛ~æ̈e]) than words like trap, cat, and rag ([æ~æ̈ɛæ̈]).[19] A similarly organized (though different-sounding) split occurs in Standard British English.
• /eɪ/, as in face, was inconsistently pronounced [e̝ː].[20]
• /oʊ/, as in goat, was inconsistently pronounced [o̝ː].[16]
• /ʌ/, as in strut, was conservative.[21]
• /ɔɪ/, as in choice, was [oɛ~oə].[22]
• /ɜːr/, as in nurse, was predominantly up-gliding, non-rhotic [ɜɪ] in the "Deep South" (all the Plantation South except North Carolina).[19]

Due to the former isolation of some regions of the Appalachian South, the Appalachian accent may be difficult for some outsiders to understand. This dialect is also rhotic, meaning speakers pronounce Rs wherever they appear in words, and sometimes when they do not (for example, "worsh" or "warsh" for "wash"). Because of the extensive length of the mountain chain, noticeable variation also exists within this subdialect.

The Southern Appalachian dialect can be heard, as its name implies, in north Georgia, north Alabama, east Tennessee, northwestern South Carolina, western North Carolina, eastern Kentucky, southwestern Virginia, western Maryland, and West Virginia. Southern Appalachian speech patterns, however, are not entirely confined to the mountain regions previously listed.

The dialect here is often thought to be a window into the past, with various claims being made that it is either a pocket of Elizabethan English which survived or the way that the Scots-Irish-origin people that make up a large fraction of the population there would have spoken. However, these are both incorrect. Though some of the distinctive words used in Appalachia have their origins in the Anglo-Scottish border region, a more realistic comparison is the way that people in North America would have spoken in the Colonial period.

Researchers have noted that the dialect retains a lot of vocabulary with roots in "Early Modern English" owing to the make-up of the early European settlers to the area.[23]

The Lowcountry, most famously centering on the cities of Charleston, South Carolina and Savannah, Georgia, once constituted its own entirely unique English dialect region. Traditionally often recognized as a Charleston accent, it included these additional features, most of which no longer exist today:[24]

• Cheer–chair merger towards [ɪə~eə].[25]
• Non-rhoticity (or r-dropping).
• A possibility of both variants of Canadian raising:
  • /aɪ/ pronounced as something like as [ɑe], but possibly [ɐɪ] before a voiceless consonant.[26]
  • /aʊ/ pronounced as [aʊ], but [əʉ~ɜʉ] before a voiceless consonant.[26]
• /eɪ/ pronounced as [ɪə~eə] in a closed syllable, [ɪː~eː] in an open syllable.[20]
• /oʊ/ pronounced as [oə~uə] in a closed syllable, [o~u] in an open syllable.[16]
• /ɔː/ pronounced as [ɔ~o].[20]
• /ɑː/ pronounced as [ɑ~ɒ],[27] with possible remnant pronunciations using even older [æ].[20]
• /ɜːr/ pronounced as [ɜ~ɞ],[27] or possibly [əɪ][19]
• /uː/ pronounced as [ʉː] or [ᵿʉ].[27]

The "Down East" Outer Banks coastal region of Carteret County, North Carolina, and adjacent Pamlico Sound, including Ocracoke and Harkers Island, are known for additional features, some of which are still spoken today by generations-long residents of its unincorporated coastal and island communities, which have largely been geographically and economically isolated from the rest of North Carolina and the South since their first settlement by English-speaking Europeans. The same is true for the very similar dialect area of the Delmarva (Delaware–Maryland–Virginia) Peninsula and neighboring islands in the Chesapeake Bay, such as Tangier and Smith Island. These two regions historically share many common pronunciation features, sometimes collectively called a High Tider (or "Hoi Toider") accent, including:

• Rhoticity (or r-fulness, like in most U.S. English, but unlike in most other older Atlantic Southern dialects)
• /aɪ/, such as the vowel in the words high tide, retaining its glide and being pronounced beginning further back in the mouth, as [ɑe] or even rounded [ɒe~ɐɒe], often stereotyped as sounding like "hoi toid," giving Pamlico Sound's residents the name "High Tiders."[28]
• /æ/ is raised to [ɛ] (so cattle sounds like kettle); /ɛ/ is raised to [e~ɪ] (so that mess sounds like miss); and, most prominently, /ɪ/ is raised to [i] (so fish sounds like feesh).[29] This mirrors the second and third stages of the Southern Vowel Shift (see under "Newer phonology"), despite this particular accent never participating in the very first stage of the shift.
• /ɔː/ pronounced as [ɔ~o], similar to modern Australian or London English.
• /aʊ/, as in loud, town, scrounge, etc., pronounced with a fronted glide as [aɵ~aø~aε].[25] Before a voiceless consonant, this same phoneme is [ɜʉ~ɜy].[26]
• /ɛər/, as in chair, square, bear, etc., as [æɚ].[25]
• Card–cord merger since at least the 1800s in the Delmarva Peninsula.[25]

The old Virginia accent was mostly spoken in the central and eastern regions of the state, excluding the Eastern Shore of Virginia on the Delmarva Peninsula.

The people of the major central (Piedmont) and eastern (Tidewater) regions of Virginia, excluding Virginia's Eastern Shore, once spoke in a way long associated with the upper or aristocratic plantation class in the Old South, often known as a Tidewater accent. Additional phonological features of this Atlantic Southern variety included:

• Non-rhoticity (or r-dropping).
• A possibility of both variants of Canadian raising:
  • /aʊ/ pronounced as [aʊ], but as [əʉ~ɜʉ] before a voiceless consonant.[26]
  • /aɪ/ pronounced as something like [aε~aæ~aə], but possibly [ɐɪ] before a voiceless consonant.[26][27]
• /ɛər/ pronounced as [æː(ə)].[27]
• /eɪ/ pronounced as [ɛ] in certain words, making bake sound like "beck", and afraid like "uh Fred."
• /uː/ pronounced as [ʉː] or [ᵿʉ].[27]
• In Tidewater Virginia particularly:
  • The "palm" words and a select few "bath" words (namely aunt, rather, and, earlier, pasture) pronounced in the mouth and often rounded as [ɒ]. This rounding, combined with non-rhoticity, potentially merge several sets of words, like maw, more, ma, and mar, into a single phoneme.[27]
  • /ɜːr/, as in bird, earth, flirt, hurt, word, dirt, etc. pronounced with weakly rhoticity as [ɜʴ] or the less common non-rhotic variant [ɜ].[19]

Southern Louisiana, as well as some of southeast Texas (Houston to Beaumont), and coastal Mississippi, feature a number of dialects influenced by other languages beyond English. Most of southern Louisiana constitutes Acadiana, dominated for hundreds of years by monolingual speakers of Cajun French,[30] which combines elements of Acadian French with other French and Spanish words. This French dialect is spoken by many of the older members of the Cajun ethnic group and is said to be dying out. A related language called Louisiana Creole also exists. The older English of Southern Louisiana did not participate in certain general older Southern English phenomena, for example lacking the Plantation South's trap–bath split and the fronting of /aʊ/.[31]

New Orleans English was likely developing in the early 1900s, in large part due to dialect influence from New York City immigrants in New Orleans.