Preparing words and grammar
A spoken word-production demonstration
In the grammar session, complete a sentence using the verb shown. In the affix sessions, learn cue-response pairs and say the paired response. Say only the answer word, as quickly and accurately as you can.
Choose a session
Space records a manual timing proxy, not a measured speech onset. On a touch device, tap the response button as you begin speaking.
After each answer, check it against the expected word. Practice trials are not included in the results.
Microphone mode requests permission only when you start. It detects sound locally; it does not record, upload, or recognize speech. Background noise can trigger it. Use a quiet room and check false triggers after each trial.
Session results stay in this page until you download them or leave. Download the CSV before closing or reloading the page.
For discussion after the demonstration
In shared blocks, items share an affix or grammatical context. In mixed blocks, the same items are regrouped so that property varies. Each item appears twice per measured condition, after practice in each block. Shared and mixed block order is randomized across sessions.
Grammar session. Each trial displays a base verb and a sentence gap, for example GO: Yesterday, I ___ home. Say only went. The verbs are GO, EAT, and BUY in three grammatical contexts:
| Context | Sentence beginning | Responses across the three verbs |
|---|---|---|
| Simple past | Yesterday, I … | went, ate, bought |
| Habitual present | Every day, I … | go, eat, buy |
| Present perfect | I have already … | gone, eaten, bought |
Every block contains all three verbs. Shared blocks use one context; mixed blocks use all three. For example, a mixed block combines GO in the simple past, EAT in the habitual present, and BUY in the present perfect. Each verb-and-sentence combination returns unchanged in both conditions. The block introduction shows the available contexts and base verbs, without listing their inflected answers. Answers appear during feedback.
This asks whether knowing the grammatical context helps prepare a response without a common affix across the three verbs. It is an extension of the classroom task, not a replication of Roelofs’s morphological comparison. Context reading, repeated sentence frames, and switching between grammatical contexts can also affect response time. A benefit would not by itself locate preparation at a particular model stage. The two contexts requiring bought remain separate items, identified by their grammatical features in the CSV.
Affix sessions. The prefix sets are unkind, unfair, untrue; reread, rewrite, repaint; and dislike, distrust, disagree. The suffix sets are careless, homeless, hopeless; joyful, helpful, playful; and kindness, darkness, softness. Here, re- means “again”; say reread as in “I will reread it.”
The preparation benefit is the median correct response time in mixed blocks minus the median in shared blocks. A positive value means faster responses in shared blocks. Errors, timeouts, practice, interrupted trials, and responses under 150 ms are excluded from that timing summary. They remain identifiable in the CSV. Error counts are reported separately.
These English affix materials are a classroom adaptation, not the materials from Roelofs’s studies. A shared affix also shares sounds and meaning. This demo has no matched sound-only control, so it cannot establish an extra morphological benefit by itself. The prefix and suffix lists also differ in frequency, stress, length, and lexical category. Compare each list with its own mixed baseline; do not treat a prefix-suffix difference as a clean test of serial position alone.
Roelofs (2000), pp. 98 to 99, reviews the stronger comparison: initial syllables that constitute morphemes versus initial syllables that do not, and initial versus non-initial morphemes in implicit preparation.