Three small words carry most of the grammar of a Filipino sentence, and none of them translates. Ang marks the participant the verb has put in focus. Ng marks a participant that is present and not in focus, which may be the thing acted on or the person doing the acting. Sa marks a place, a direction or a recipient. Learn those three as a system and the sentence opens; learn them as translations and it stays shut.
The trap is that each of them has a plausible English gloss. Ang looks like the, ng looks like of, sa looks like to. Two of those three will produce a wrong reading of an ordinary sentence within a week.
Six sentences that show the focus system working
- Kumain ako ng mansanas, I ate an apple, with the eater in focus. The affix -um- inside the verb announces that the focused noun is the doer.
- Kinain ko ang mansanas, I ate the apple, with the apple in focus. Same event, different affix, and the pronoun changes from ako to ko because it is no longer the focused one.
- ang, ng and sa, the three markers: focus, non focus, and location or direction. Every noun in a sentence carries one of them.
- mag- and -um-, two doer focus affixes with a real difference: mag- tends to a deliberate or ongoing action and -um- to a single completed one.
- may and mayroon, there is, with the second used when something follows it directly and the first when a noun does.
- Ako si Ana, I am Ana. Si marks a personal name in focus, a separate marker from ang, and names always take it.
Why the doer is sometimes marked like a possessive
When the verb puts the object in focus, the object takes ang and the person doing the action takes ng. To an English eye that sentence reads as the fish of the man, and the learner concludes that they have misparsed something. They have not. The marker is not saying possession, it is saying not in focus, and the same form does both jobs.
This is where the case markers and the affix on the verb turn out to be one mechanism rather than two. The affix announces which role is in focus; the marker in front of each noun confirms it. Read them together and the sentence is unambiguous even with the words in an unusual order. Read only the nouns and it is a puzzle.
Each marker has two more forms that have to be learned alongside it: one set for personal names and one set of pronouns. There is no way to skip these, because a sentence with a name in it uses the name series and nothing else, and pronouns are in the majority of sentences anyone says.
Aspect, and the syllable that gets repeated
Filipino verbs do not mark tense. They mark aspect: whether the action is finished, ongoing, or not yet begun. Three forms, built from the same root, and the ongoing and not yet begun forms both repeat the first syllable of the root. That repetition is not decoration; it is the grammar, and it is the part learners drop first when they speed up.
Two systems are therefore stacked on the same word. The affix says which participant is in focus, the shape of the stem says which aspect, and both have to be produced at once. A form with the right affix and no repeated syllable is a different sentence, not a sloppier one.
One root, three aspects
What changes in the word as the action moves
These are states of the same verb rather than steps in a method. The repeated syllable is the part that carries two of them.
- 01The bare rootNot usable in a sentence at all, and the form a vocabulary list gives you
- 02FinishedThe affix appears, and nothing is repeated
- 03Going on nowThe affix, plus the first syllable of the root said twice
- 04Not yet begunThe repeated syllable again, with the affix in its other position
A speaker who leaves out the repetition has not spoken carelessly; they have said a different one of these four frames. It is the first thing to go when a learner speeds up, and the easiest to hear once you know to listen for it.
Where it stalls: reading ng as of
Around the third or fourth month a learner can build actor focus sentences accurately and stops dead on any sentence where the doer is marked with ng. The symptom is specific: they parse it as a possessive, get a nonsense reading, and blame their vocabulary.
It persists because it is invisible. Nothing is wrong with the words they know, and the sentence is short. What is wrong is a mapping they made in week one and never revisited, and mappings like that do not surface on their own.
The exit is a habit rather than a rule. When a sentence resists, find the verb's affix first, decide which role it puts in focus, and only then read the nouns. Doing that deliberately twenty times converts it into the thing you do automatically, and this is one of the few grammar problems in this pillar where a fortnight of conscious drilling genuinely ends it. How long that fortnight sits inside the whole project is hours to fluency, by language and by starting point.
Two things worth using
The first is a marker table you write out yourself: three columns for the three markers, three rows for ordinary nouns, personal names and pronouns, and nine cells filled in from sentences you have actually met. It replaced the glossary line that translates ang as the, which is the source of the error above, and it costs an afternoon and nothing else.
The second is a language model asked to label a sentence rather than translate it: mark each participant, name the affix, say which role is in focus. As of 2026 the free tiers do this reliably for Filipino, and it replaced working backwards by hand from a conjugation chart, which requires already knowing the system you are trying to learn. Check its labels against your own before you accept them and it becomes an exercise rather than an answer key.
The mistake: learning roots and hoping
A vocabulary list of Filipino roots looks efficient and produces sentences you cannot say. The root is not a word; it is what a word is built from, and the built form carries both the focus and the aspect. Storing the bare root leaves both of those to be worked out under time pressure.
Store a whole form instead, inside a short sentence, with its markers visible. It is three times the material per card and roughly a tenth of the confusion later, which is a good trade and a specific case of making cards that teach rather than cards that pile up. Meeting those forms often enough for the pattern to become obvious is material just past your level, and why that is the sweet spot.
The next question is which words go into the slots, and the answer is unusual here, because a large part of the everyday vocabulary arrived from two colonial languages and kept its shape, in Spanish and English loans sitting on a Tagalog frame. And the small words that attach themselves to the first word of a clause have an order of their own, in everyday exchanges, and the English mixed into them.