Researching a Language Sound System
Chapter Objectives
By the end of this chapter, you will be able to:
- Define a language variety and scope precisely enough to support a phonetic profile.
- Distinguish primary evidence from secondary description.
- Evaluate the quality of a phonetic source.
- Identify conflicting analyses and explain why they exist.
- Create a provisional source and example ledger.
- Formulate answerable phonetic questions for a chosen variety.
- Design a manageable research plan for the capstone profile.
An Observable Problem
Every chapter from 6 through 28 supplied the language examples for you, already verified, already labeled with a reference variety, already narrowed to the phenomenon under discussion. Chapters 30 through 32 will ask you to do the opposite: choose a language variety yourself, gather your own evidence, and build a documented phonetic profile that another careful reader could check.
That shift creates a new kind of problem, one that has nothing to do with articulation, acoustics, or notation. Suppose a learner decides to profile "Spanish." A search turns up a pronunciation dictionary describing one national standard, a grammar describing a different regional variety, a language-learning video with confident but unsourced claims, and a peer-reviewed article that flatly contradicts the video on where a particular fricative is produced. Nothing in Chapters 1 through 28 tells this learner which source to trust, how to reconcile the contradiction, or whether "Spanish" was ever a well-defined research target in the first place.
This chapter asks three questions:
- What makes a language variety a well-defined research target rather than a generic label?
- How does an analyst judge whether a source's phonetic claims deserve confidence?
- What has to be planned and documented before inventory construction can begin responsibly?
Key Terms
- Language variety: a specific, named form of a language, defined by region, community, era, or standard, precise enough that a phonetic claim about it can be checked.
- Primary evidence: a recording, an original transcription, or a directly observed production by an identified speaker of the target variety.
- Secondary description: a source's account or analysis of primary evidence, such as a grammar, dictionary, or textbook, produced by someone other than the speaker.
- Source hierarchy: a ranked approach to evidence quality that favors peer-reviewed and official sources over informal or unsourced ones without treating rank as an automatic guarantee of accuracy.
- Analytical alternative: a differing, defensible account of the same phonetic or phonological fact, arising from different data, theory, or reference variety.
- Source ledger: a running record connecting each claim or example in a profile to its source, variety, and level of confidence.
- Research scope: the bounded set of questions, evidence, and variety that a capstone profile commits to addressing.
Defining a Language and Variety
Purpose: apply the reference-variety policy from Chapters 5 through 23 to the larger task of choosing an entire language variety as a research target, before any inventory work begins.
Every chapter so far has required a named reference variety whenever variation could affect an example. Chapter 29 extends that requirement to the scale of an entire profile. "French," "Arabic," and "Chinese" are not phonetic research targets by themselves. Each label covers multiple standards, regional forms, and, in some cases, distinct languages grouped under one popular name. A workable research scope narrows the label to something an identified source can actually describe: a named standard, a specific city or regional form, a particular generation of speakers, or a clearly bounded community.
Narrowing a variety is not busywork. It determines which sources are even relevant. A pronunciation dictionary for one national standard cannot responsibly supply the vowel inventory of an unrelated regional variety, even if both share a language name. Chapter 18's caution against "regular French" or "neutral Spanish" applies here at full scale: a profile that never states its variety cannot be verified, compared, or corrected.
Some widely used language names cover several mutually unintelligible varieties, and some named varieties cross what popular usage treats as separate languages. Neither situation is unusual, and neither should be resolved by the learner's convenience. Choose the variety label that matches the actual sources available, state it explicitly, and let that label govern every claim that follows.
Error: Naming a broad language family or a country as the "variety" and treating any source that mentions the language name as equally relevant.
Correction: Name a specific standard, region, or community, and reject sources that describe a different variety under the same broad label, even when locating a well-matched source takes more effort.
Primary Evidence and Secondary Description
Purpose: give learners a working distinction between evidence and description of evidence, extending the phonetic-versus-phonological distinction from Chapter 5 to the research process itself.
Primary evidence is as close as a researcher can get to the sound itself: a recording of an identified speaker, an original phonetic transcription made from that recording, or, where available to the learner, direct observation of a speaker's production. Secondary description is an account of that evidence produced by someone else: a grammar's summary of the consonant inventory, a dictionary's pronunciation key, a textbook's simplified explanation, or a popular article's claim about how a sound is produced.
Both kinds of evidence are useful, and a capstone profile at this level will rely mainly on carefully selected secondary description, since Chapter 32's boundaries do not require fieldwork or native-speaker access. The distinction matters because secondary description always involves a layer of analysis and simplification that primary evidence does not. A grammar's phoneme count reflects the author's phonological decisions as well as the raw acoustic and articulatory facts; a recording does not. Treating a secondary source's convenient summary as if it were the sound itself repeats the phone-versus-phoneme confusion Chapter 5 first warned against, now at the level of an entire source rather than a single example.
| Feature | Primary evidence | Secondary description |
|---|---|---|
| Closeness to the sound itself | Direct: a recording, an original transcription, or observed production | Indirect: an account written by someone analyzing the sound |
| Typical form | Audio, video, an original field transcription | Grammars, dictionaries, textbooks, articles summarizing findings |
| Main risk | Limited scope: one speaker, one context, one recording session | Analytical bias, oversimplification, or outdated claims layered onto the evidence |
| Role in a capstone profile at this level | Used when available; not required by Chapter 32's boundaries | The main evidence base for most learners, selected using the source hierarchy in Section 29.3 |
Evaluating Source Quality
Purpose: teach a practical, ranked approach to source evaluation that a learner can apply without specialist training, while avoiding the false idea that rank alone settles every question.
Not every account of a language's sound system deserves equal trust. A useful default ranking, from strongest to weakest for non-obvious technical claims, moves through: official materials from a recognized phonetic authority; peer-reviewed phonetic and phonological research and its associated journals; scholarly grammars and language descriptions; established pronunciation dictionaries and reference works; reputable university-level textbooks; and, at the bottom, general web pages, unsourced pronunciation sites, and machine-generated transcriptions, which are not sufficient evidence on their own for a non-obvious claim.
Rank is a starting heuristic, not a verdict. A peer-reviewed article can still describe a variety different from the one a learner has chosen; a well-regarded grammar can still be decades out of date on a rapidly changing sound pattern. Evaluate every source on four practical questions: does it name the variety and speaker population precisely enough to match the chosen scope; does it disclose its own evidence base, such as recordings, fieldwork, or prior literature; does it distinguish its own analysis from simple description; and does it agree, disagree, or stay silent where other credible sources speak.
A high-ranked source that fails the variety-match check is less useful than a lower-ranked source that matches the chosen scope precisely. Apply the ranking in Section 29.3 as a first filter, then apply the four practical questions to every source that passes it. A source's prestige is not a substitute for confirming that it actually describes the variety under study.
Conflicting Analyses
Purpose: normalize disagreement among credible sources and give learners a way to represent it honestly, extending the analytical-alternatives principle already used for transcription in Chapters 24 and 26.
Two credible sources can disagree about the same variety for reasons that have nothing to do with either source being wrong. They may rely on different speaker samples, use different theoretical frameworks, describe the variety at different points in time, or draw the boundary between phonemic and phonetic claims differently. A profile that silently picks one source and ignores the conflict misrepresents the state of knowledge; a profile that reports the disagreement, briefly and clearly, is more accurate and more useful.
When two sources disagree, first check whether the disagreement is genuine or only apparent: a difference in transcription convention or notation system, of the kind addressed in Chapters 9 and 25, can look like a substantive disagreement without being one. If the disagreement is genuine, name both positions, state what appears to motivate each, and select the position the profile will adopt, with a brief, honest note explaining the choice. This is the same posture Chapter 24 required for transcription-level decisions, applied now to whole-source disagreements.
A learner who has spent the whole course learning that spoken language varies systematically should expect the scholarly record to vary too. Treating every disagreement among credible sources as a mistake to be resolved in favor of one "correct" answer misrepresents how phonetic and phonological research actually works. State disagreements plainly rather than hiding them behind a single confident claim.
Citation, Rights, and Ethical Representation
Purpose: connect the course's evidence-and-source discipline to citation practice and to the identity and variation ethics established in earlier chapters, now applied to an entire profile's source base.
Every claim, example, or transcription drawn from a source belongs to that source and must be attributed. Record enough information about each source to cite it consistently: author or organization, title, publication context, date, and the specific claim or page the profile relies on. Where a source's material is protected by copyright or a specific license, confirm that quoting, adapting, or reproducing it for the profile is permitted, and prefer summarizing a source's findings in your own words to reproducing its examples verbatim.
The variation and identity policy from Chapter 18 governs source selection as directly as it governs example writing. Do not select or discard sources based on which regional or social variety they favor without a stated, defensible reason. Do not describe a variety as more correct, more standard, or more authentic than another variety of the same language merely because more sources happen to describe it. Where a chosen variety belongs to a community with its own preferred names and self-descriptions, use those names rather than an outside label imposed on the community.
A phonetic profile is a description of how a community of speakers uses sound to communicate, not a specimen collected for the researcher's convenience. Favor sources produced by, or in direct collaboration with, members of the speaker community when they are available and credible. Where only outside description exists, say so plainly in the profile's limitations rather than presenting the outside account as if it were uncontested.
Building a Source and Example Ledger
Purpose: teach the concrete documentation habit that will support Chapters 30 through 32, turning the abstract evidence policy of this chapter into a reusable working record.
A source and example ledger is a running table, kept outside the learner-facing profile, that connects every claim or example to its origin. Recording this information as research proceeds, rather than reconstructing it afterward, prevents the common failure of an otherwise careful profile that can no longer say where a specific claim came from.
| Claim or example | Language and variety | Source | Evidence type | Confidence | Notes |
|---|---|---|---|---|---|
| Consonant inventory includes a distinct palatal nasal | Named variety, to be confirmed | Scholarly grammar, cited edition | Secondary description | High | Confirm against a second grammar before finalizing |
| Word-final obstruents are regularly devoiced | Named variety, to be confirmed | Peer-reviewed article | Secondary description, based on recorded data | High | Author states data source and speaker count |
| A specific vowel merger is spreading among younger speakers | Named variety, to be confirmed | Recent sociolinguistic study | Secondary description | Moderate | Single study; treat as an emerging pattern, not an established fact |
Every row states the claim, the variety it is claimed to hold for, the source, whether that source offers primary evidence or secondary description, an honest confidence level, and any notes needed to remember why that confidence level was assigned. This structure previews the fuller source ledger required by the Chapter 32 capstone and gives Chapters 30 and 31 a ready foundation to build on.
Error: Collecting interesting examples first and planning to track sources later, once the profile has taken shape.
Correction: Log the source alongside every claim the moment it is gathered. A claim with no recorded source becomes unusable evidence, no matter how accurate it later turns out to be.
Formulating Answerable Questions
Purpose: teach learners to convert a broad interest in a language's sound system into specific, source-answerable questions, setting the research agenda that Chapters 30 and 31 will fill in.
A broad ambition such as "understand the sound system of this language" cannot be checked against evidence directly. An answerable question names a specific phonetic or phonological property, states the variety it concerns, and can plausibly be answered from an identifiable kind of source. Compare the two levels below.
A manageable capstone plan usually rests on somewhere between six and twelve answerable questions, distributed across the areas Chapter 32 will require: consonant and vowel inventories, major allophones, orthography-sound relationships, syllable structure and phonotactics, relevant prosodic systems, connected-speech processes, and variation. Questions that cannot be matched to any available source should be narrowed, replaced, or explicitly listed as a limitation rather than answered by guesswork.
Scoping Different Kinds of Varieties
A well-documented national standard with an established descriptive tradition supports a narrow, confident scope, since multiple credible sources are likely to exist. A lesser-documented regional or minority variety may support only a handful of sources, in which case the profile's honest limitation is that it rests on a small evidence base rather than a large one. A variety undergoing rapid, well-studied change, such as a documented vowel shift, may need a scope that explicitly names a generation or time period alongside the region. The research method taught in this chapter applies to all three situations; what changes is how much confidence the available evidence can support, and how clearly that confidence level must be stated in the final profile.
Research-Planning Exercises
These exercises work from described research situations rather than live source searching, so that planning and evaluation skills can be practiced and checked before you begin locating your own sources for Chapter 30.
Exercise A — Revising a Variety Label
A learner proposes to profile "German." Explain why this label is not yet a usable research scope, and propose two different ways it could be narrowed into a workable one.
Exercise B — Matching Claims to Evidence
Three claims are given: a claim from a peer-reviewed article about a stop-voicing contrast, a claim from an undated pronunciation website about the same contrast, and a claim from a scholarly grammar about a related vowel pattern. Rank the three claims by evidentiary strength and justify the ranking using the four practical questions from Section 29.3.
Exercise C — Handling a Genuine Disagreement
Two scholarly grammars of the same named variety disagree about whether a particular vowel pair is contrastive or allophonic. Describe the steps you would take before deciding how to represent this disagreement in a capstone profile.
Exercise D — Writing Answerable Questions
Turn the broad interest "I want to understand the vowels of this language" into three specific, answerable research questions appropriate for a capstone plan.
Common Mistakes
Error: Treating a source's high rank on the source hierarchy as proof that it applies to the chosen variety.
Correction: Confirm the variety match first. A lower-ranked source that actually describes the chosen variety is more useful than a prestigious source describing a different one.
Error: Resolving every disagreement between credible sources by silently choosing the source that seems more familiar or more convenient.
Correction: Investigate why the sources disagree, and represent a genuine disagreement honestly in the profile rather than concealing it behind one confident claim.
Error: Writing broad research ambitions instead of answerable questions, leaving no clear way to know when a question has been sufficiently answered.
Correction: Rewrite each ambition as a specific, variety-bound question that names the phenomenon and the kind of source expected to answer it.
Chapter Review
Part A — Scope and Evidence
1. Explain what makes a language variety a usable research target, and give one example of a label that is not yet usable.
2. Distinguish primary evidence from secondary description and explain why most capstone profiles at this level rely mainly on the second.
Part B — Source Evaluation
1. List the source hierarchy from strongest to weakest for non-obvious technical claims.
2. State the four practical questions used to evaluate a source beyond its rank.
Part C — Planning
1. Explain what a source and example ledger records and why it should be kept as research proceeds rather than afterward.
2. Rewrite the ambition "learn how this language sounds" as one answerable research question.
Cumulative Review
1. From Chapter 18: Explain how the reference-variety policy first applied to individual examples now applies to the scope of an entire capstone profile.
2. From Chapter 28: Explain how this chapter's treatment of evidence strength parallels Chapter 28's distinction among supported, plausible, and unestablished acoustic claims.
3. Preview for Chapter 30: Predict how the source ledger and answerable questions built in this chapter will be used directly when constructing a consonant and vowel inventory.
Self-Check
Before moving to Chapter 30, confirm that you can do the following without looking at your notes:
- State a language and variety precisely enough for another reader to identify it.
- Distinguish primary evidence from secondary description in a given source.
- Apply the source hierarchy and the four practical evaluation questions to a real or described source.
- Explain what to do when two credible sources disagree.
- Build a provisional source and example ledger entry for a single claim.
- Write at least three specific, answerable research questions for a chosen variety.
If any of these tasks are difficult, review the relevant section before reading Chapter 30. Chapter 30 uses the variety, sources, and questions established here to construct an evidence-based consonant and vowel inventory.
Answer Key (click to expand)
Research-Planning Exercises — Exercise A
"German" is not yet a usable research scope because it covers multiple national and regional standards with documented phonetic differences, so a source describing one standard cannot be assumed to describe another under the same broad label. Two workable narrowings include naming a specific national standard, such as the standard associated with a particular country's public broadcasting or reference dictionaries, or naming a specific regional variety with its own descriptive literature, clearly distinguished from the national standard.
Research-Planning Exercises — Exercise B
The peer-reviewed article ranks strongest, since it sits near the top of the source hierarchy and, presumably, states its evidence base and speaker population. The scholarly grammar's claim ranks next, as a respected secondary description likely to disclose its sources even though it addresses a different, related pattern rather than the same contrast. The undated pronunciation website ranks weakest, since it sits at the bottom of the hierarchy and, without a stated evidence base or publication context, cannot be confirmed to satisfy the variety-match, evidence-base, or consistency checks from Section 29.3.
Research-Planning Exercises — Exercise C
First confirm the disagreement is genuine rather than a notational difference, by checking whether both grammars use compatible transcription conventions for the vowel pair in question. If the disagreement is genuine, investigate what might explain it, such as different speaker samples, different eras of data collection, or different phonological frameworks. Then decide which position the profile will adopt, with a brief stated reason, and note the alternative analysis explicitly rather than presenting only one grammar's conclusion as settled fact.
Research-Planning Exercises — Exercise D
Acceptable answerable questions include, for example: "How many contrastive vowel qualities does the named variety have, according to a scholarly grammar or peer-reviewed source?"; "Does the named variety show a length distinction among its vowels, and what evidence supports that claim?"; and "Are there documented allophonic variants of any vowel in the named variety, and in what contexts do they occur?" Each names the variety, a specific vowel-related phenomenon, and the kind of source expected to answer it.
Chapter Review — Part A
1. A usable language variety names a specific standard, region, generation, or community precisely enough that a source's claims can be checked against it; a broad language-family or country-level label, such as "German" used without further qualification, is not yet usable because it can refer to multiple distinct standards. 2. Primary evidence is a recording, an original transcription, or directly observed production; secondary description is an account of that evidence written by someone else, such as a grammar or dictionary. Most capstone profiles at this level rely mainly on secondary description because Chapter 32's boundaries do not require fieldwork or native-speaker access, so carefully evaluated existing descriptions form the realistic evidence base.
Chapter Review — Part B
1. From strongest to weakest: official materials from a recognized phonetic authority; peer-reviewed phonetic and phonological research; scholarly grammars and language descriptions; established pronunciation dictionaries and reference works; reputable university-level textbooks; and general web pages, unsourced pronunciation sites, and machine-generated transcriptions. 2. Does the source name the variety and speaker population precisely enough to match the chosen scope; does it disclose its own evidence base; does it distinguish its own analysis from simple description; and does it agree, disagree, or stay silent where other credible sources speak.
Chapter Review — Part C
1. A source and example ledger records each claim or example, the language and variety it concerns, its source, whether that source offers primary evidence or secondary description, an honest confidence level, and supporting notes; it should be kept as research proceeds because a claim collected without its source becomes unusable evidence later, and reconstructing sources afterward is unreliable. 2. An acceptable answerable question is, for example: "What consonant and vowel inventory does a scholarly grammar or peer-reviewed study report for the named variety, and do multiple sources agree?"
Cumulative Review
1. Chapter 18 required every language-specific example to identify a reference variety whenever variation could affect it; this chapter applies the same discipline to an entire capstone profile, requiring the whole project, not just individual examples, to commit to one clearly defined variety before evidence gathering begins. 2. Chapter 28 distinguished acoustic claims that are supported, merely plausible, or not established by a display alone; this chapter applies the same three-way distinction to sources, separating claims a source clearly supports from claims it only suggests and claims it cannot establish without further, independent evidence. 3. Chapter 30 will likely draw directly on the variety definition, evaluated sources, and answerable questions built here, using the source ledger as the evidentiary basis for each consonant and vowel placed in the inventory and using the research questions to determine which gaps in the inventory still require further sourcing.
Looking Ahead
What You Can Now Do
You can now define a language variety and scope precisely enough to support a phonetic profile; distinguish primary evidence from secondary description; evaluate source quality using a practical hierarchy and four evaluation questions; identify and represent conflicting analyses honestly; build a provisional source and example ledger; and formulate answerable phonetic questions. Together with the acoustic evidence discipline of Chapter 28, you now have the full evidentiary foundation the mapping chapters require.