Trademark, work and proof: the gap Brazil's music registry still leaves open

1. Starting point: what the law requires

Article 122 of the Brazilian IP Act conditions registration on a visually perceptible sign. That is why Brazil, unlike the United States and the European Union, has historically not admitted sound or scent marks.

This has an underexplored consequence. When a gesture is registered — a hand forming a sign — it is not the gesture that becomes protected. It is its representation. The score, not the melody.

The distinction resolves apparent conflicts. Someone performing the gesture is not reproducing the figurative mark: they are performing an act. Trademark use would arise only if the image were applied, as a distinctive sign, to a related product or service.

Art. 122 — visually perceptible distinctive sign

2. Why a sound mark would not solve the sampling problem

The intuitive proposal is to register samples as sound marks, creating a state mechanism to support Brazilian artists against international copying. The intuition about the problem is right. The instrument is what falls short.

A trademark protects a sign that distinguishes goods or services in commerce. A phone-maker's identification tune is a mark because it identifies that manufacturer — not because it is music. If someone sampled that tune inside a song, there would be no trademark infringement: there is no use as a distinctive sign for related goods.

Copyright protects the work itself — composition, melody, phonogram. Unauthorised sampling falls under Law 9,610/98.

Registering a sample as a trademark would protect against someone using that sound to identify a soft drink. It would do nothing against another musician incorporating it into a track.

Where sound marks genuinely work in music

In the producer's signature. Beat tags — the short timbre opening funk, trap and authored production tracks — operate exactly as marks: they identify the origin of a production service, not the work being sold. There the distinctive function exists. It is a real market, and still unstructured in Brazil.

3. The functionality test

One test has been proposed with precision: remove the sample and check whether the work still does what it set out to do. If it does not, the sample is essential.

The test is correct — and it decides against trademark registration, not for it. Article 124(XXI) bars signs inseparable from a technical or functional effect. The rationale is economic: trademarks renew indefinitely, and perpetual exclusivity over what is indispensable to functioning would lock up the market.

But the argument proves something else, and proves it well: if removal breaks the work, the excerpt is a substantial part of it — and substantial parts are already protected by copyright, with no registration required.

The right does not need to be created. What fails is proof and enforcement.

Art. 124(XXI) — sign inseparable from technical or functional effect

4. The examples, precisely

The thesis is usually supported by cases of international copying of Brazilian music. They are worth distinguishing, because they do not all point the same way.

• Gotye — “Somebody That I Used to Know” (2011). The track samples “Seville” by Luiz Bonfá, and Bonfá is credited as co-author. The use was licensed. This is not appropriation: it is the system working as intended.
• Rod Stewart — “Da Ya Think I'm Sexy?” (1978). Reproduces a melodic line from “Taj Mahal” by Jorge Ben Jor, initially uncredited. Stewart acknowledged the similarity and the matter was settled. This is the case that illustrates the problem.

Using the first as an example of misappropriation weakens the thesis before a technical reader. The second sustains it.

5. The right instrument: certification marks

The stated objective — the State signalling that it stands with the artist — does not require a product mark. It requires a certification mark (Art. 123(II)) or a collective mark (Art. 123(III)).

A certification mark attests conformity with technical standards and is held by an entity without direct commercial interest in the activity. It does not protect the sound. It certifies the work: declared authorship, verifiable date, file integrity.

• The seal is registrable under existing law;
• The work remains protected by copyright, as it always was;
• And what was missing — technical proof of priority — becomes the content of the certification.

The artist gains no monopoly over a timbre. They gain an enforceable certificate: this work existed on this date, with this content, and has not been altered.

Art. 123(II) — certification mark · Art. 123(III) — collective mark

Limits of reach

A Brazilian seal does not bind a foreign court. What crosses borders is not the stamp — it is the evidence. The Berne Convention already guarantees automatic protection in signatory countries; what was always missing was the evidentiary basis to exercise it.

6. The chain of proof in derivative works

When a work is built by altering a pre-existing excerpt, the dispute is never “copied or not”. It is: what was added, and when.

Without a record of the intermediate stages, the author arrives at litigation holding the final product and nothing showing the path to it. With a certified chain — original excerpt, treatment applied, date of each layer — authorship of the transformation becomes visible. Which is precisely what copyright protects in a derivative work.

Hence a technical requirement of the system: registration must reach the stages, not only the result.

7. Practical effect: criminal and economic

Articles 184 and 186 of the Brazilian Criminal Code already criminalise copyright infringement, and search and seizure is already available. Certification does not change the offence — it changes the investigation. Police and prosecutors need documented materiality to act quickly. Today the author arrives with an allegation; with a certificate, they arrive with pre-constituted proof.

Economically the effect is more direct. Trademarks communicate registration; copyright communicates nothing. The ® symbol tells the market there is a holder and that examination occurred. No equivalent symbol exists for a sample.

Without an identifiable unit, with clear ownership and verifiable history, licensing cannot be structured at scale — not for lack of demand, but for lack of a transactable object.

8. The structural gap

Article 124(XVII) bars registering as a trademark a protected intellectual work belonging to a third party, without the author's authorisation. It should be read precisely: it does not bar authors from registering what they created — it protects against appropriation by others.

And here lies the core of this analysis. That conflict analysis is not performed. The INPI — like other offices — does not cross-check trademark applications against catalogues of copyrighted works. Not by administrative choice, but because no comparable base exists: copyrighted works are not indexed, not normalised, and carry no technical identifier allowing automated comparison.

The provision exists on paper and is unenforceable in practice.

The law need not change to justify the system. It is enough to show that the law already requires an analysis nobody is able to perform.

Art. 124(XVII) — protected third-party work, without authorisation

9. What the absence of registration produced

Authorship does not depend on registration. The principle is sound, and it is also where the practical problem begins.

A right that is not documented is a right that communicates nothing. The buyer does not know whom to buy from, the user does not know whom to ask, and the copier meets no sign warning them off. The work circulates with no visible owner — and the market treats what has no apparent owner as having none.

The change in tools made this worse. Producing a mark once meant hiring a draughtsman; patent drawings were done by hand. Today an image is generated by prompt, or lifted from the first search result. Creating and copying came to cost the same: nothing. And once the cost is equal, the sense that someone else's labour is involved disappears.

The effect is generational, and it is sharpest in Latin America, where enforcement is thinner and the habit of licensing never took hold. What formed is an audience that does not recognise the right — not out of bad faith, but because it has never seen the right represented anywhere.

This is where the symbol matters more than it appears to. The ® creates no right at all: it communicates that an owner exists and that an examination took place. Nothing equivalent exists for a musical work. There is nothing to point at, and so there is nothing to respect.

Right in existence · evidence absent · recognition nonexistent

10. Advertising recognised what the register still has not

There is a use the market practised well before any legal discussion: the fragment of a song placed in a commercial is not there as a work. It is there as a recognition signal — the role a logo performs, executed through a different sense.

The typical Brazilian case is the advertising parody: a familiar song has its lyrics swapped to associate the product with the advertiser, and the listener identifies the brand before processing the words. Advertising use of this kind is done under licence — and that is precisely the point. The fragment has a price, has an owner and is transacted. Nobody arrived at the technique in theory: advertising found it by trial, because it noticed that sound reaches the consumer by a route the visual mark cannot take.

An even fuller pattern appears when a campaign remixes a familiar song and alters the hook so it becomes phonetically close to the brand name, keeping the original word at points. The stated purpose is usually to fix the brand in the audience's memory. The whole operation is on display there — a fragment of a work, reworked, performing source identification without ceasing to be a work. And because it is a rework, it is also the case that demands the chain of proof described in section 6: without a record of the layers, authorship of the transformation cannot be shown.

The difference is sensory, and worth stating precisely: a figurative mark depends on the consumer looking. Sound does not request attention — it captures it. That is the economic reason sound signatures became an advertising asset long before they became a registrable category.

Where the argument meets its limit

Before the limit, what holds. Trademark registration gives rise to a right of its own: patrimonial, autonomous from the copyright already subsisting in the same creation, transferable and renewable for equal and successive periods, with no final term. Cumulative protection is not an anomaly — it is the ordinary regime of every logo, which is a work and a mark at once.

There is even an asymmetry favouring that thesis, and it deserves naming. Art. 132(IV) bars the owner from preventing the citation of their mark in a speech, a scientific or literary work or any other publication, provided there is no commercial connotation and no prejudice to distinctiveness. No reciprocal provision exists: copyright law does not open the work to trademark use, and Art. 124(XVII) still requires the author's authorisation. The mark is permeable to the work; the work is not permeable to the mark.

Read a contrario, however, the same provision narrows the ground of the preceding section. An advertising piece carries commercial connotation by definition, and there the owner may prevent. Song parody in a commercial lives in that band — not in the freedom to cite.

And the limit proper is a different one: a sign cannot be the mark of the very thing it is. A melody is registrable in the classes where it indicates origin — recordings, production, entertainment services. What cannot be registered is music as a mark for music: there it is the product, not the indicator of it, and perpetual exclusivity over what is indispensable falls under the Art. 124(XXI) bar already discussed.

There is also what registration does not reach. An author's moral rights are inalienable and unwaivable, and no trademark registration dispenses with attribution when the work is performed as a work. A mark adds a new patrimonial layer over the same creation — it does not replace the earlier one.

What remains registrable is everything around it, and that is not little: the producer's signature, the label's sting, the timbre that opens the tracks of a catalogue — signs identifying who produced, not what is sold. It is a real market, already operating, and still unstructured in Brazil.

And this is where certification returns as the correct instrument. To license a fragment, the fragment must first exist as a unit: with defined limits, clear title and a verifiable date. Without that, every negotiation restarts from zero and most simply never happen — not for refusal, but for want of an object.

Art. 132(IV) — citation of a mark in a work, without commercial connotation · Art. 124(XVII) — third-party work, without authorisation

11. Conclusion

Protection is not what is missing. The ability to demonstrate it is.

Authorship arises upon creation and requires no registration — but that is theory. Whoever cannot prove when they created holds no enforceable right: they hold an allegation.

The certificate does not create the right. It shifts the burden. Without it, the author proves. With it, the other side disproves.

It is the same architecture IPmorrow developed for document integrity — compare, map and publicly prove that a result exists, bears a certain date and has not been altered — applied to a catalogue nobody has organised.

RM.IP · Rodrigo Monteiro — Industrial Property Agent, INPI 2249
IPmorrow — creator of the BIPPO System

← Back