6.2 Mapping Objects, Actions, and Properties

Turning a scene into parts

Many visual-symbolic drills can be read as miniature sentences. An object behaves like a noun, an action like a verb, and a property like an adjective or adverb. This is a study analogy, not an official DLAB claim, but it works because language aptitude depends on mapping form to meaning in fresh systems, the exact thing the DLAB measures.

Start with objects. If a picture of a cup is labeled "mav" and a key is labeled "ren," object identity may set the whole label. But if a red cup is "lom mav" and a blue cup is "tir mav," then "mav" likely means cup while "lom" and "tir" mark colors. You need contrast: one picture is rarely enough to fix a rule.

Actions add a layer. Suppose a dot moving upward is "sa," a dot moving downward is "lo," and a square moving upward is also "sa." The action (or direction) is stronger than object identity. If a later item shows a triangle moving upward, the likely action label is still "sa." That is transfer across object type, and recognizing it quickly is worth real points because Category IV-track candidates need 110+ and cannot afford to relearn each item from scratch.

Minimal-pair comparisons

A minimal pair is a pair of examples differing in exactly one important feature. It is powerful because it isolates the rule. If everything changes at once you may still solve the item, but your inference is weaker. Hunt for the pair that changes least.

Properties are classic distractors. Bright color, an unusual shape, or large size draws the eye even when the label actually tracks number or position. Train the question: did that feature vary with the label? If every red object has a different label, red is probably not the main clue. If every small object shares a label piece, size becomes a strong candidate.

Actions encode more than motion

An arrow is rarely just "moves." A right arrow could mean moves right, gives to the right, points right, or transforms into the right-hand object. Read the captions and answer choices carefully. If the examples show a ball becoming a square, the arrow may mark transformation, not movement. Common DLAB-style traps reward candidates who notice the result or the affected object rather than only the direction of travel.

Common traps

• English-order assumption: never assume the invented label follows English word order. The marker may come first, last, or be infixed.
• Single-example certainty: committing to a rule from one picture is the fastest way to miss a transfer item.
• Salience bias: choosing the loudest feature (bright red, biggest shape) instead of the feature correlated with the label.

For self-study, draw four pictures with the same object and different properties, then four with different objects and the same property. Assign invented labels consistently. After a short delay, try to infer your own rule from the table. This rehearses the exact mental move the DLAB rewards: extracting the feature that survives across changes while ignoring the features that do not.

Order of operations within a label

Object-action-property systems often combine markers, and the order of markers is itself a clue. Suppose "big red cup" is "tev lom mav" while "small red cup" is "nu lom mav." The size word leads, color sits in the middle, and the object word ends. If a new item is "big blue key," you can predict "tev tir ren": size-color-object, slotting the new color and object into the established frame. Many candidates infer the right meanings but place the markers in the wrong order and lose the point.

Always check whether the system is consistent about position, because invented grammars frequently use fixed word order to carry information the way real case-marking languages do.

A second order trap is the boundary marker. Some practice systems separate a property cluster from the object with a fixed particle, for example "lom-ka mav" where "-ka" closes the description before the noun. If you ignore the particle you may misread where the property block ends and the object begins. Treat punctuation-like symbols as grammatical, not decorative.

Why this is worth drilling

Given roughly two hours for 126 questions, your per-item budget is under one minute. Object-action-property fluency is what makes that budget survivable: once you can name the feature inventory in five seconds and slot markers into a known frame, the easy two-thirds of items become near-automatic, leaving margin for the genuinely hard transfer problems. The candidates who score above 110 are rarely the ones who reason fastest on any single hard item; they are the ones who never waste time on the easy ones.