A child looks at an empty cup, holds it toward a caregiver, and says, “More juice.” That brief message contains far more clinical information than its length suggests. The child is expressing an idea about additional juice, organizing that idea into a two-word combination, and directing the message toward another person for a specific purpose. An SLP then examines what the child means, how the message is structured, and what the child is trying to accomplish. Those three questions correspond with the language framework introduced by Lois Bloom and Margaret Lahey in Language Development and Language Disorders in 1978. Their work connected language development with the assessment and treatment of children with language disorders and included detailed attention to language transcription and semantic content categories.
The powerful Bloom and Lahey model gives SLPs a way to go beyond broad statements such as “expressive language is delayed” or “the student has difficulty with pragmatics.” Instead, the clinician can examine which part of the language system appears to be limiting communication. Is the person having difficulty representing the idea itself? Is the idea understood but difficult to encode through morphology or syntax? Is the linguistic structure available, but the person is not providing the information the listener needs?
What Are Content, Form, and Use
The model is most useful when the SLP avoids treating these as three isolated categories. Real communication typically draws on all three at once.
Content Is More Than Vocabulary
Content is related to how the ideas and relationships are represented in language. Vocabulary is part of semantics, but content extends well beyond whether a child knows the names of objects. Compare the following utterances:
“Dog.”
“Dog sleeping.”
“Dog under table.”
“The dog hid because the thunder frightened him.”
The first identifies a referent. The second adds an action or state. The third represents a location relationship. The fourth expresses a causal relationship between two events.
In each case, the speaker is doing more than adding vocabulary. The speaker is representing increasingly complex relationships among concepts.
Bloom and Lahey's framework has been used to examine semantic categories involving concepts such as existence, recurrence, negation, possession, action, location, state, quantity, time, and causality. More recent research reinforces the importance of examining relationship among linguistic concepts. A 2023 study of 82 Mandarin-speaking preschoolers examined semantic content categories in naturalistic language samples and found developmental differences in the number and type of relationships children expressed. The study also identified linguistic and cultural influences on acquisition patterns, which is an important reminder that developmental sequences described in English should not automatically be applied to every language.
This distinction is especially important during academic language tasks. Suppose a student watches ice melt and says, “The water came.” An SLP could focus immediately on sentence structure, but a content analysis first asks whether the student understands the conceptual relationship among heat, melting, and the change from solid to liquid. If the underlying causal relationship is not represented accurately, teaching the word because alone may produce a grammatical sentence with incorrect content.
A student could say, “The ice melted because it was cold.” The syntax is acceptable. The explanation is not.
Content and form therefore need to be scored separately when the treatment target requires both.
Form: How Language Is Organized
Form relates to structures through which language is represented; it identifies phonology, morphology, and syntax as its main components.
Phonology is how the sound system of a language is organized. Morphology is referred to the units that carry grammatical or lexical information, such as plural -s, past tense -ed, third-person singular -s, possessives, prefixes, and suffixes. Syntax is how words and phrases are organized into sentences.
It's important to draw a distinction because a grammar target, a phonological pattern, and a motor speech difficulty are not interchangeable problems. The analysis needs to identify which linguistic structure is being examined rather than treating every spoken error as a problem with “form.”
Think of a child who says, “Yesterday she walk home.” The child may understand the temporal concept and correctly refer to an event in the past, but the required past-tense morphology is absent. The semantic content and grammatical form therefore show different levels of development.
Now compare that with a child who says, “The girl was running home because she forgot her umbrella,” but produces several speech-sound distortions. The sentence may be grammatically sophisticated even though the spoken production is difficult to understand. Intelligibility and grammatical form should not automatically be combined into one score.
The framework also applies outside conventional speech. An AAC user can express complex content, combine symbols into grammatical structures, and use those messages for different purposes. If a person has difficulty producing multi-symbol messages, the clinician needs to determine whether the limitation reflects language learning, restricted vocabulary on the system, motor access, navigation demands, partner behavior, or some combination of those factors.
Use: What Communication Accomplishes
Use relates to pragmatics, but it should not be attributed to conversational etiquette.
A person may communicate to request, refuse, comment, direct, ask, answer, explain, negotiate, repair, advocate, share an experience, or provide information for another person. The central question is what the function of that particular message is.
This is one reason eye contact, body posture, or other conventional social behaviors should not become universal measures of successful language use. Social norms differ across individuals, cultures, families, and communication communities. ASHA explicitly recognizes that behaviors such as eye contact, gesture, proximity, tone, and conversational style vary considerably.
A more useful pragmatic question is whether the message accomplishes the communicator's purpose.
Imagine that a student says, “Put it there.” The communication partner responds, “I'm not sure which one you mean.” The student's next response provides information about communication repair. They may point, repeat the sentence, name the object, add a descriptor, use AAC, or abandon the message.
Interaction Among Content, Form, and Use
Suppose the student is asked why a tower fell and responds:
“It fell. The table.”
The SLP can ask whether the student understands that someone bumped the table and caused the tower to fall. That is a question about content.
The clinician can also examine whether the student has the morphology and syntax needed to express the relationship through a structure such as, “The tower fell because he bumped the table.” That is a question about form.
Finally, the clinician can ask whether the explanation gives the listener enough information to understand what happened. That is a question about use.
These dimensions can diverge. A child might understand the cause but fail to produce the targeted grammatical structure. Another might produce a perfectly grammatical causal clause but identify the wrong cause. A third might express the correct event and use appropriate grammar but rely on pronouns whose referents the listener cannot identify.
One combined “language accuracy” percentage would conceal those differences.
Applying the Model During Assessment
A practical assessment does not need to begin by deciding which Bloom and Lahey category is weak. It is usually more useful to begin with the functional communication problem.
Suppose a teacher reports:
“The student has difficulty explaining what happened during classroom activities.”
Several different problems could cause a concern. The student may omit important content, such as the root of an event. They may understand the relationship but have difficulty constructing the sentence needed to explain it. They may produce the sentence but fail to provide enough background information for a listener who was not present.
Those possibilities suggest different assessment probes and different treatment goals.
A useful clinical sequence is to identify the participation problem, observe a situation where it occurs, describe the content, form, and use demands, select a primary target, choose a measure that matches that target, and then check whether performance transfers to different materials, partners, or environments.
Language Samples and the Bloom and Lahey Framework
Language sampling is particularly compatible with this model because connected communication allows the SLP to observe semantics, morphology, syntax, and pragmatics within context rather than as isolated test items.
The sampling activity, however, strongly influences what language is elicited.
A preschool construction-play sample may produce actions, locations, requests, comments, and short descriptions. A science explanation may produce temporal and causal relationships. A personal narrative may require event sequencing, tense, reference, and enough information for an unfamiliar listener. A peer activity may elicit negotiation, questions, and repair. An AAC-supported activity may reveal how easily the person can comment, refuse, direct, or build symbol combinations.
Hollis and Petersen studied 190 school-age children using brief and longer narrative elicitation tasks and found differences across several measures depending on the sampling procedure. Changes in elicitation can therefore complicate interpretation of progress.
If the initial sample is a highly structured story retell and the later sample is an open personal narrative, a change in MLU or sentence complexity may partly reflect the task rather than language ability.
How Long Should a Language Sample Be?
The sample needs to be long enough to provide sufficient opportunities for the behavior being measured. A broad measure such as utterance length may require less sampling than a rare grammatical structure or pragmatic event.
Muir and colleagues studied 21 deaf and hard-of-hearing preschoolers and found that several broad measures derived from three-minute samples showed strong reliability relative to longer samples. The measures included MLU and word-productivity variables. That study supports the possible use of brief samples for selected measures in that population, but it does not establish that three minutes is sufficient for every grammatical target, pragmatic function, population, or intervention outcome.
If the SLP is measuring communication repair and no communication breakdown occurs during the sample, the clinician has no denominator for judging repair. If the target is third-person singular -s and the sample contains no obligatory contexts, lack of production cannot be interpreted as lack of ability.
The sampling procedure should therefore be designed around the question being asked.
Measure the Skill Being Treated
This may be one of the most useful findings in the contemporary language-sampling literature.
Gallagher and Hoover examined six preschool children treated for third-person singular -s. Measures specific to the treatment target captured improvement that broad language measures such as MLU and type-token ratio did not consistently reflect.
The sample was small, so it does not establish one preferred outcome measure for every language intervention. It does illustrate an important principle.
If therapy targets causal sentence structures, measure causal sentence structures.
If therapy targets communication repair, measure repair.
If therapy targets semantic relationships, measure those relationships.
MLU can provide useful information about utterance length, but longer utterances do not automatically demonstrate improvement in every language domain.
A student can produce a longer sentence that contains inaccurate content, poor cohesion, or ineffective listener information. Broad measures should therefore complement rather than replace target-specific measures.
From Language Sample to Treatment Goal
Consider a classroom activity in which a tower falls after a student bumps the table.
The SLP asks:
“Why did the tower fall?”
The student responds:
“It fell. The table.”
Before deciding on a goal, the SLP examines the three components. Does the student understand the causal relationship? Can the student construct a causal clause? Does the message give the listener enough information?
A proposed treatment goal could be:
Given a pictured or observed event, the student will explain an outcome and its cause using a targeted causal sentence structure with the level of support and criterion established from baseline data.
The clinician can then score the response across several dimensions.
This separation becomes useful when responses differ.
“Because he bumped it” may contain accurate content and the target form.
“The cup tipped. Water came out” may communicate the cause accurately without using the targeted because clause.
“Because it was cold” may be grammatically accurate but factually incorrect for the observed event.
A single score would obscure those distinctions.
Applying Content, Form, and Use During Preschool Play
The same approach works with younger children.
During construction play, a child says:
“Block there.”
The message expresses a location relationship, uses a two-word combination, and may function as a direction to the adult.
The clinician can respond according to the current treatment target. If the goal addresses syntax, the SLP might model, “Put the block there.” If the goal is for deeper semantic relationships, the adult might model, “The block goes under the bridge.”
The model should respond to the child's actual communicative intention rather than expanding every short utterance in the same way.
Applying the Model During Everyday Routines
During dressing, a child holds a shoe toward a caregiver and says:
“Shoe.”
The utterance could function as a label, a request for help, or an attempt to direct the caregiver's attention. The interaction determines the likely function.
If the shoe is stuck, the clinician might acknowledge the child's existing message and model, “Help me” or “Shoe stuck.”
Naturalistic intervention has an evidence base separate from Bloom and Lahey's framework. In a randomized trial of caregiver-implemented Enhanced Milieu Teaching, caregiver strategy use improved and children showed an advantage in receptive language, while broad expressive-language outcomes did not show the same intervention effect.
The Bloom and Lahey model can help identify what the clinician wants to observe and target. It does not establish treatment efficacy by itself.
Applying the Framework to AAC
AAC provides another useful example of why content, form, and use should be considered together.
Suppose a child selects:
MORE + CAR
during play.
The content represents recurrence or additional activity. The form is a two-symbol combination. The use is a request for continuation.
The clinician should then broaden the observation. Can the person comment, refuse, ask for help, direct another person, share information, or repair a misunderstanding? ASHA's AAC guidance emphasizes assessment of language, access, communication environments, and communication partners rather than measuring requesting alone.
The AAC system itself can also affect form. A person may know the grammatical relationship they want to express but lack convenient access to the required words or grammatical markers. Navigation demands may reduce message length. Motor access may slow production. A communication partner may interrupt before the message is completed.
An AAC language sample therefore needs to consider both linguistic ability and access conditions.
.png)
.png)