CONCEPTUAL BOUNDARIES
Concept and application: five distinctions that keep thought open
An application may open a concept, but should not silently exhaust it.
A question that comes before selection
The Atlas explains what each entry contributes and why it merits a place. Before those editorial decisions comes a question about the reach of an explanation: when does a linguistic application illuminate a broader concept, and when does it quietly take its place? An entry's function and its admission rationale remain separate matters. The distinctions developed here do not, by themselves, justify admitting a term.
A useful analysis brings something within reach. We can examine how a speaker identifies a source of information, how a sentence presents a conclusion as possible or necessary, or how a narrative invites us to orient ourselves within an imagined scene. Each approach reveals something. None warrants the assumption that everything worth understanding about perception, knowledge or imagination has now been captured.
The difficulty begins when a precise tool acquires the authority of a complete account. The language of a study may move from 'we investigate this aspect through these expressions' to 'the phenomenon consists of these expressions'. Five distinctions help readers notice that movement. They mark differences in scope while keeping the connections available for examination.
Five distinctions and the connections they preserve
| Distinction | What the analysis addresses | What a reduction would leave out | A defensible connection |
|---|---|---|---|
| Perception is not evidentiality. | Evidentiality concerns linguistic ways of marking an information source. Perception includes sensory and embodied experience. | Experience that is unspoken, prelinguistic or not organised as the source of a claim. | Source marking provides one approach to how experience and evidence are represented in language. |
| Knowledge is not epistemic modality. | Epistemic modality concerns what may or must be the case relative to an information base. Knowledge raises further philosophical questions. | Truth, justification, error and the distinction between knowing and believing. | Modal language shows how a claim is assessed against information or assumptions; it does not establish that its speaker knows. |
| Imagination is not deixis at the phantasma. | Bühler's deixis am Phantasma concerns pointing within an imagined or absent field. Imagination has a wider range. | Imagined sounds, moods, images and possibilities without a linguistically organised centre of orientation. | Deictic expressions can give an imagined scene a situated perspective. |
| Ontology is not merely taxonomy. | A taxonomy groups entities into classes. Ontology also examines existence, kinds of being, identity and dependence. | The question of what makes a category appropriate and whether its members exist independently of the classification. | A taxonomy can express ontological commitments that remain open to scrutiny. |
| Representation is not automatically semiosis. | Representation is a broad term for standing for or depicting something. Peircean semiosis concerns mediation through sign, object and interpretant. | Differences among theories of depiction, mental representation and sign action. | Peirce's triadic account offers a theoretical framework whose suitability should be explained rather than assumed. |
1. Perception and evidentiality
Consider three reports: 'I can see smoke above the roof'; 'There must be a fire inside'; 'The caretaker says there is a fire inside'. They lead a listener towards a similar concern, but they make different grounds for the claim available: observation, inference and another person's report. A source can be identified explicitly, expressed through a construction, or supplied by context.
Linguistic descriptions distinguish systems with obligatory grammatical source marking from lexical and contextual ways of identifying a source. In a broad discussion of evidential strategies, all three reports are relevant. In a narrower grammatical account, they should not simply be treated as instances of the same category. Nor should inferential 'must' automatically be assigned only an evidential analysis: epistemic and evidential functions can interact.[1]
The distinction matters because perception also takes place before a report is made. Feeling pressure on a hand, watching a moving shadow or hearing a change in pitch need not result in an assertion with a marked information source. The linguistic account illuminates what becomes communicatively available; it does not encompass the entire experience.
2. Knowledge and epistemic modality
English gives us 'She has arrived', 'She may have arrived' and 'She must have arrived'. The second leaves an option open; the third presents arrival as required by the relevant information. Neither expression settles whether the information is reliable. A convincing inference can still start from a mistaken premise.[2]
Here the English version should do its own explanatory work. German wohl has no single equivalent that preserves every use and nuance. Translating it mechanically as probably would conceal the importance of context. English may and must offer a clearer contrast for this discussion, but each also has other uses: 'You may enter' grants permission, and 'You must leave' can express an obligation. The form alone does not identify an epistemic reading.
Knowledge is therefore more than a degree of verbal confidence. A person may speak firmly while being wrong, or cautiously while possessing good grounds. Philosophy disputes how knowledge should be analysed; linguistic markers help us examine an assertion's presentation and commitments without resolving that dispute.
3. Imagination and pointing within an imagined field
Imagine the timbre of a cello, a colour you have never seen in that room, or a future in which a difficult conversation ends differently. None requires a narrative 'here' from which left and right become intelligible. Imagination can work through qualities, possibilities and feelings as well as through spatial scenes.
Now read: 'You are standing at the gate. To your left is the garden. Over there, beyond the wall, the path begins.' The reader is offered a location from which the directions make sense. The pointing does not depend on sharing the speaker's present physical surroundings. It builds an imagined field of orientation.
Bühler discusses deixis am Phantasma in his Sprachtheorie (1934). The German label is retained here because an English explanatory phrase such as 'pointing within an imagined field' makes the mechanism clearer than an unexplained literal rendering. His English Theory of Language provides a further route into that terminology.[3] In each example, we still need to show that a deictic field is being organised. Language stimulating an image is not sufficient evidence for this particular operation.
4. Ontology and taxonomy
A library may classify a work as a novel, a historical document or a political essay. A biological taxonomy groups organisms according to selected relationships or characteristics. Such classifications make comparison and retrieval possible. They also select what counts as a significant resemblance.
Ontology asks questions that remain open after the arrangement has been made. Are the kinds discovered or constructed? What allows a thing to remain the same thing through change? Are processes or relationships as fundamental as objects? Competing accounts can use similar lists while disagreeing about what those lists commit us to.[4]
A technical use of ontology may deliberately narrow its remit to a particular domain. Such a convention should be made explicit. Comparing the labels in a classification does not establish that the larger philosophical questions have been answered. The Atlas should state which sense and level of analysis are at work.
5. Representation and semiosis
A map represents a region. A warning light stands for a condition that calls for attention. Peirce's theory gives us a way to examine how the relation becomes effective through interpretation: a sign relates to an object through an interpretant. The interpretant should not be confused with the interpreting person; it concerns the mediating interpretation or effect through which the sign operates.[5]
That account brings movement into view. The warning light becomes a reason to check a machine because a relation is taken up and produces a further understanding or response. The account has explanatory force because its triadic mediation is specified, not because the word semiosis has been attached to any instance of depiction.
Representation has a wider and varied theoretical history. An image, a mental state and a formal model need not be described by precisely the same account. Calling them all Peircean semiosis is a theoretical choice. Making that choice visible helps readers examine it.
How a method comes to stand for its object
The shift can happen through a sequence of reasonable decisions. A broad question becomes a manageable research question; observable expressions become evidence; the selected evidence becomes an operational definition. The trouble comes when the operational definition travels back into general explanation without its original limits.
A variable can be useful precisely because it excludes much of the phenomenon. Counting the occurrence of source expressions may answer a question about a corpus. The resulting count is not a measurement of everything involved in human perception. As the method becomes familiar, its exclusions can become harder to notice.
- Begin with a broad concept and choose an aspect to investigate.
- Identify expressions, constructions or responses that provide observable evidence.
- Define how those observations will count within the investigation.
- Check whether the general explanation has inherited a narrower definition without saying so.
- Restore the cases and questions the method leaves outside its scope.
A routine for keeping the connection precise
The Atlas can make a connection useful by specifying its reach. 'Evidentiality is one way language makes an information source available' tells a reader what to look for. 'Perception is evidentiality' prematurely closes the question. The first statement offers a researchable relation; the second replaces a broader subject with one analytical approach.
- Which level is being described: a capacity, a philosophical concept, a linguistic form, a communicative action or an analytical result?
- Which relation is claimed: part, realisation, example, operationalisation, analogy or theoretical account?
- Which relevant cases remain outside the chosen approach?
- Can a counter-example separate the broader concept from the application?
- Which theory and terminological convention does the explanation assume?
What the continuation of a sentence makes salient
'X is not Y' corrects an identification, but can be heard as denying any connection. A relation needs to follow the distinction. Even the continuation changes the route of thought: it may put experience, linguistic form or an unresolved limit in the foreground.
The following continuations offer different emphases. They are interpretive prompts rather than interchangeable definitions. English does not have to repeat the same 'not … but' construction in every cell; what matters is that the movement of emphasis remains visible.
| Starting point | The broader level | A linguistic approach | An open limit |
|---|---|---|---|
| Perception is not evidentiality … | Perception involves embodied contact with a world before any assertion is made. | A report can identify an experience as seen, heard, inferred or received from another person. | Source marking reveals only part of what is experienced. |
| Knowledge is not epistemic modality … | Knowledge raises questions about truth, reasons and possible error. | Epistemic language relates a proposition to information and assumptions. | The presentation of certainty does not establish philosophical warrant. |
| Imagination is not deixis am Phantasma … | Imagination reaches absent, possible and not-yet-actual situations. | Deixis can organise a particular imagined field around a centre of orientation. | An imagined sound or mood need not have that centre. |
| Ontology is not merely taxonomy … | Ontology asks what kinds of being and dependence an account recognises. | A classification makes some commitments available for examination. | The existence of a classification does not settle whether its categories are fundamental. |
| Representation is not automatically semiosis … | Representation covers several ways of standing for or depicting something. | A Peircean analysis specifies mediation among sign, object and interpretant. | The fit of that account has to be argued for in the particular case. |
Sentence structure helps select what the reader treats as decisive. A continuation can broaden a concept or subtly narrow it. Noticing the selection is itself a movement of thought: we hear the sentence as an arrangement of attention, not merely a container of information.
Separating the concept from its favoured example
- A person feels pain and makes no assertion: perception occurs without linguistic source marking.
- Someone says 'The parcel may arrive today' while lacking knowledge of whether it will: epistemic possibility is expressed without establishing knowledge.
- A listener imagines a melody without locating it in a scene: imagination operates without pointing within a deictic field.
- Two taxonomies can arrange the same material under different criteria: the arrangement alone does not decide which kinds are ontologically fundamental.
- An image depicts something: a Peircean account still needs to specify the sign relation and its mediation through an interpretant.
A counter-test is a way of preserving explanatory reach. It shows both what an application reveals and why its favourite example should not become the definition of the whole concept.
What this changes in reading the Atlas
The Atlas gives readers tools for following linguistic movements. Those tools become more useful when a relation is named: a profile may describe a linguistic realisation, a limited aspect, a theoretical framework or a neighbouring concept. A link makes that relation available for inspection; it does not make the two endpoints synonymous.
The five distinctions propose a working rule: an application may open a concept, but should not silently exhaust it. Readers can follow an explanation, use it on a case and notice where the larger question remains. That combination supports understanding without creating an impression of completeness that the analysis cannot justify.
The next part explains the four entry functions and the three admission grounds. Those labels concern an entry's editorial role and the reason for its place in the Atlas. This chapter concerns a different question: how far an application explains, and where a broader concept must remain open.
Sources and conceptual reference points
- Alexandra Y. Aikhenvald: Evidentiality (2004) · Alexandra Y. Aikhenvald: Evidentiality, Annual Review of Anthropology
- Epistemic Modality, Internet Encyclopedia of Philosophy
- Karl Bühler: Sprachtheorie (1934), bibliographic record · Karl Bühler: Theory of Language, John Benjamins
- Metaphysics, Stanford Encyclopedia of Philosophy · Categories, Stanford Encyclopedia of Philosophy
- Peirce’s Theory of Signs, Stanford Encyclopedia of Philosophy · Charles S. Peirce: MS 634.23–28 (1909), Arisbe transcription
The five distinctions and the checking routine are editorial proposals for the Atlas. They are not quotations or attributed doctrinal statements from the authors cited.