A photo shows what is visible at one moment. It rarely proves who the people are, what happened before, or why an action is taking place. Unsupported inference begins when a response turns plausible guesses into facts.
The official Duolingo English Test guide is the source of truth for the current question types and instructions. Whatever photo task appears in your current guide, the language principle is stable: ground the response in the image, distinguish observation from interpretation, and do not replace description with a memorized story.
Practice the broader task vocabulary through the Lyra Practice Duolingo English Test vocabulary hub and DET Write About the Photo vocabulary guide.
Use three evidence zones
Sort a possible sentence into one of three zones:
- visible: directly observable objects, positions, colors, actions, or setting features;
- supported: a cautious interpretation based on several visible clues;
- invented: identity, motive, emotion, cause, or event with no visual support.
Imagine a photo with two adults in reflective vests beside a road. One holds a clipboard; cones surround a shallow hole.
- visible: “Two adults in reflective vests are standing beside a section of road marked by cones.”
- supported: “They appear to be inspecting or discussing the marked area.”
- invented: “Two city engineers are investigating a dangerous gas leak after residents complained.”
The third sentence adds employer, profession, hazard, and history. None is visually established.
Mistake 1: assigning exact identities
Unsupported: “A doctor is explaining a diagnosis to her patient.”
Evidence-led: “A person in clinical clothing is speaking to another adult across a desk, and both are looking at the same document.”
If a badge clearly says Doctor, you can report the label. If the room merely looks clinical, use appears to be a clinic or describe the visible equipment.
Identity words are tempting shortcuts: teacher, manager, scientist, tourist, and family. Replace them with visible roles when uncertain:
- a person presenting to a group;
- an adult holding a map;
- two people examining equipment;
- several people seated around a table.
This remains specific without pretending to know more than the photo shows.
Mistake 2: reading emotion from a neutral expression
Unsupported: “The woman is angry because the service is terrible.”
Evidence-led: “The woman is leaning toward the counter and gesturing with one hand while speaking to the employee.”
A clear smile may support appears cheerful. Tears may support appears upset. But emotion labels are interpretations, and their causes are another inference again.
Use cautious forms sparingly:
- appears focused;
- seems relaxed;
- may be surprised;
- their posture suggests close attention.
A hedge does not rescue a baseless claim. “She is seemingly furious about being dismissed” remains invented if the image shows only a raised hand.
Mistake 3: inventing purpose
Unsupported: “The group has gathered to plan a charity event.”
Evidence-led: “Five people are gathered around a table covered with papers, and one person is pointing to a large calendar.”
Supported interpretation: “The arrangement suggests a planning discussion.”
The photo may support planning through the calendar and shared attention. It does not support charity without a label or recognizable material.
Separate action from purpose:
- action: writing on a board;
- possible purpose: appears to be explaining a process;
- invented purpose: is training new employees for tomorrow’s safety audit.
Understanding is only the first step.
Lyra Practice helps you retrieve and use high-value workplace expressions in realistic situations until they feel natural.
Start a practice session →Mistake 4: creating a before-and-after story
A still image does not normally establish sequence.
Unsupported: “The cyclist fell, injured his knee, and is now waiting for an ambulance.”
Evidence-led: “A bicycle is lying beside the path, and a person is seated nearby while another adult crouches beside them.”
You may describe arrangements that suggest assistance: “The second person appears to be checking whether the seated person is all right.” Do not narrate the fall unless the image itself contains a sequence or explicit evidence.
Mistake 5: mistaking location clues for certainty
Overcertain: “This is a university library in London.”
Grounded: “The setting appears to be a library or study area, with shelves behind several people working at long tables.”
Architecture and furniture can support a setting category. They rarely identify a city or institution. If a readable sign names the place, report it. Otherwise, stay with indoor study space, market, transit station, or another defensible category.
Mistake 6: adding judgment instead of description
Judgmental: “The careless workers created a dangerous mess.”
Descriptive: “Several tools and open boxes are spread across the floor near the workers.”
Careless, dangerous, inefficient, and successful are evaluations. Describe the feature that creates the impression. If a cable visibly crosses a walkway, say so; do not assume an accident will occur.
Build detail through spatial relationships
When you avoid stories, you still need enough content. Spatial language makes visible details cohere:
- in the foreground/background;
- to the left/right of;
- beside, behind, opposite, or between;
- surrounded by;
- facing toward/away from;
- running along the edge of.
Thin: “There are people, trees, and a building.”
Developed: “In the foreground, two people are walking along a tree-lined path. Behind them, a low brick building stretches across the image, while several bicycles are parked near its entrance.”
All the added information is visible and organized.
Choose exact action verbs
Replace doing something with an observable action:
- leaning over a table;
- pouring liquid into a container;
- pointing toward a screen;
- carrying a stack of boxes;
- arranging items on a shelf;
- waiting beside a platform.
Do not make the verb stronger than the evidence. A person touching a machine may be adjusting it, but not necessarily repairing it. Someone holding a book may be looking at it, but not demonstrably studying for an examination.
Worked response: outdoor scene
Imagined photo: a narrow path crosses a wetland. Three people with backpacks walk away from the camera. A sign stands near the path, and dark clouds appear above distant hills.
Unsupported version:
“Three tourists are escaping a dangerous storm after ignoring a warning about floods.”
Evidence-led version:
“The photo shows three people with backpacks walking along a raised path through a wetland. A sign stands beside the path, although its text is not readable, and low hills are visible in the distance. Dark clouds cover much of the sky, so the weather appears unsettled. The group is moving away from the camera toward the hills.”
The revision supplies objects, location, movement, and a cautious weather interpretation. It does not invent danger, nationality, or a past decision.
Worked response: indoor scene
Imagined photo: two adults in aprons stand behind a counter. One places pastries in a display case; a customer looks through the glass.
Unsupported version:
“The bakery owners are preparing expensive desserts for a disappointed customer.”
Evidence-led version:
“Two adults wearing aprons are working behind a food counter. One is arranging pastries inside a glass display case, while a customer stands on the other side and looks at the selection. Shelves of bread are visible in the background, so the setting appears to be a bakery or café.”
The photo supports the setting category through several clues, but it does not establish ownership, price, or emotion.
A product-led practice routine
Open the Lyra Practice DET photo vocabulary guide and use each visual prompt in three passes. First, list only visible nouns and actions. Second, connect them with spatial relationships. Third, add no more than one cautious interpretation and underline its evidence. Compare the three passes: every final sentence should trace back to the image.
Then use the DET writing vocabulary by question type guide to keep photo-description language separate from extended argument language.
Final self-check
- Can I point to visible evidence for every noun and action?
- Did I assign a profession, relationship, or location without a label?
- Did I invent what happened before or what will happen next?
- Have I confused a facial expression with a confirmed emotion?
- Did I add a purpose or cause that the photo cannot show?
- Is my cautious inference based on more than one visible clue?
- Did I organize details spatially instead of listing them?
- Are my action verbs observable rather than speculative?
- Did I describe the current photo rather than use a memorized scene?
Accurate description can still be rich. Build richness through precise objects, actions, and relationships—not unsupported biography.
Sources
Related
- Duolingo English Test Vocabulary Practice
- DET Write About the Photo vocabulary
- DET photo spatial relationships
- DET Speaking vocabulary by question type
- DET Writing vocabulary by question type
The Duolingo English Test is a trademark of Duolingo. Lyra Practice is not affiliated with or endorsed by Duolingo.