Same Sound, Different Spelling: Phonetic Similarity in Trademark Searches

Phonetic similarity means that two trademarks spelled differently sound the same or very nearly the same. Because search tools match letters, searching for a single spelling will not find marks that sound identical; you need to search, one by one, every spelling that could produce the sound of your mark. In a phonetic similarity trademark search, sound is an independent dimension that oppositions weigh alongside visual and conceptual similarity.

Below you will find the family of variants you need to search, a 20-row example table built with fictitious names, root and wildcard search techniques, and a checklist for sorting the results you find.

What Is Phonetic Similarity and Why Does It Matter in a Search?

Aural (phonetic) similarity refers to how alike two signs sound when pronounced. According to the Trademark Examination Guidelines (Marka İnceleme Kılavuzu) of TÜRKPATENT (Turkish Patent and Trademark Office), this comparison looks at the number and order of syllables, stress and shared syllables. Trademarks also travel through word of mouth, phone orders, radio ads and voice search, and in those channels the consumer never sees the spelling.

To understand where the risk arises, you need to separate the two examination channels. According to the Guidelines, in the Office's ex officio examination under Article 5(1)(ç) of Industrial Property Law No. 6769 (SMK), aural similarity alone is generally not enough for a finding of "indistinguishable similarity" unless there is a high degree of visual similarity. In an opposition under Article 6(1), by contrast, aural similarity is weighed separately as one of the three dimensions of likelihood of confusion.

The upshot is this: an application that sounds the same as an earlier mark but is spelled differently can pass the Office's examination, be published in the Official Trademark Bulletin and then face an opposition from the owner of the earlier mark. Before filing, you are the only one who can see that risk. We cover the general criteria for similarity in our article on the likelihood of confusion, and whether those criteria can be reduced to a percentage in our article on trademark similarity percentage.

Why Is Searching a Single Spelling Not Enough?

Because databases match characters, not sounds. TÜRKPATENT's search screen offers "equals" (ile eşit), "starts with" (ile başlayan) and "contains" (içinde geçen) options, and all three work on letter strings. A search for KAMEVO will not return a record for CAMEVO, even though the two are pronounced the same. Phonetic similarity does not surface in trademark databases by itself; you can only catch it by querying the right variants one at a time.

Webx's free trademark search tool also searches for the exact text you type and lists the records found with their status, classes and image; it is up to you to enter the pronunciation variants. Unless you know how a tool handles Turkish characters, search spellings such as ŞELVANO and SELVANO separately.

We explain the general search method, built on exact, core-word and class layers, in our guide on how to do a trademark search. This article focuses on the second layer of that method: pronunciation variants.

The Family of Variants You Need to Search

The family of variants means every spelling that someone who hears your mark might write down, or that a competitor might choose. It falls into five groups.

Turkish and foreign letter equivalents

Letters missing from the Turkish alphabet, and letters unique to Turkish, produce different spellings of the same sound:

• k / c / q: in foreign spellings the "k" sound is often written with a "c" or a "q".
• ş / s / sh, ç / c / ch: without Turkish characters, "ş" (the "sh" sound) is often written as "s" or "sh", and "ç" (the "ch" sound) as "c" or "ch".
• ğ: the Turkish "soft g" lengthens the vowel before it; in foreign spellings it is rendered as "g" or dropped altogether.
• ı / i, ö / o, ü / u: spellings without Turkish characters, plus the German-style equivalents "oe" and "ue".
• w / v, x / ks, j / c: Turkish readers pronounce "w" as "v" and say "x" as "ks", and the Turkish letter "c" gives the sound of the English "j".

The Guidelines deal separately with how these differences play out in ex officio examination. Long words that differ only by "x" and "ks" are, as a rule, considered indistinguishably similar; in short words, or where the difference sits at the beginning of the word, the outcome can change. With "K" and "C", the pronunciation stays the same but the appearance can differ, so aural similarity alone is generally not enough for an ex officio refusal. At the opposition stage, aural similarity is weighed separately in both cases.

Double letters, dropped letters and repeated letters

A doubled letter often does not change the pronunciation, and sounds that are weak in speech, such as "h", can disappear from the spelling. According to the Guidelines, in longer words a repeated letter in the middle or at the end generally does not create enough difference to take the signs out of indistinguishable similarity in ex officio examination. In short words, however, a single repeated letter can change perception.

Suffixes, endings and spelling format

Suffixes create a new-looking word while keeping the core of the mark. Typical forms to search include:

• Turkish suffixes: affectionate suffixes such as "-cım" and "-cik" and possessive suffixes such as "-m" and "-miz". According to the Guidelines, the effect of a possessive suffix depends on the distinctiveness of the word it is added to and on whether it changes the overall perception of that word.
• Foreign endings: "-ist", "-ix", "-tech". The Guidelines treat signs with the same root that differ only by the endings "-tech", "-tek" and "-tec" as, in principle, indistinguishably similar. Adding an ending such as "-tech" to a word, on the other hand, does not as a rule lead to that result in ex officio examination; that does not mean there is no risk on opposition.
• Joined, separate and hyphenated spelling: according to the Guidelines, punctuation such as hyphens, full stops or commas placed between words does not, as a rule, create a difference. The "&" sign, however, is read as "ve" or "and" and can therefore change the aural perception of short marks.

Numeral, letter and symbol equivalents

Numerals are read both as numbers and in place of similar-sounding words: "4" can be said as "dört" (Turkish for four), "four" or "for", and "2" as "iki" (Turkish for two), "two" or "to". According to the Guidelines, an "@" used inside a word is, as a general rule, read as the letter "a". The same number written in figures and in words is not, on its own, enough for an ex officio refusal if there is no visual similarity; on opposition, however, it comes into play because the pronunciation stays the same. We cover whether marks made of letters and numbers can be registered in our article on letter and number trademarks.

Foreign pronunciation, plural and singular

According to the Guidelines, the average consumer is assumed not to know a foreign word and to read it according to Turkish rules; for simple words in a widely known language such as English that have entered everyday use, the correct foreign pronunciation is also taken into account. You therefore need to search a foreign word under spellings that reflect both its Turkish reading and its original reading. Whether a foreign word can be registered as a trademark at all is a separate question, which we address in our article on registering foreign-word trademarks.

For the English plural "-s", the Guidelines' general rule is clear: outside short words, this ending does not remove indistinguishable similarity. Search the singular and plural forms, together with the forms carrying the Turkish plural suffix "-ler/-lar".

Example Variant Table

All the trademark names in the table below are fictitious examples invented for this article. Each row shows, for one type of variant, the name searched and the additional spellings that should be searched:

The table is a starting template: if your mark contains more than one type of variant, combine the types as well. For a name containing both "ş" and "x", for example, you also need to try spellings in which both equivalents change at the same time.

How Do You Run a Root and Wildcard Search?

A root search means identifying the unchanging core of a mark and retrieving every record that contains that core. A wildcard is a character that stands in for unknown letters in a query. Used together, they let you cover dozens of variants with fewer searches.

The "starts with" and "contains" options on TÜRKPATENT's search screen serve as root searches. According to the help page of TMview, run by the EUIPO, the asterisk (*) stands for zero or more characters and the question mark (?) for exactly one character. TMview also offers a fuzzy search based on the Levenshtein algorithm.

1. Break the mark into syllables and identify the most distinctive one.
2. Remove the problem letters (such as c/k, ş/s or x/ks) and write down the core that stays constant; for KAMEVO, that is "AMEVO".
3. Search the core with the "contains" option.
4. In databases that support wildcards, replace the problem letter with "?" (?AMEVO); if the number of letters may vary, use "*" (*AMEVO).
5. Run a "starts with" search on the opening syllable; the Guidelines stress that consumers pay more attention to the beginning of a word.
6. In tools that offer fuzzy search, run the same name through that option as well.
7. Narrow the growing list of results to your own classes under the Nice Classification.

Wildcard searching comes at a price: the number of results grows fast. Start with a narrow wildcard ("?" for a single letter) and widen it only if needed. Note the exact query you used each time, so that you can repeat the same search word for word just before filing.

How Do You Sort the Results?

A variant search produces a long list, and not every record carries the same weight. Prioritize records that genuinely sound close to yours and sit in the same or a related field. The Guidelines' criteria for aural comparison bring the following questions to the fore when sorting:

• Is the opening syllable the same? The beginning of a word is more decisive in both visual and aural comparison.
• Do the number, order and stress of syllables match? Shared syllables and the same rhythm are the strongest indicators of similarity.
• Is the difference in an unstressed syllable? An unstressed syllable added to the middle or end of a word generally does not change its aural character.
• Does the match rely on a regional pronunciation? For Turkish words, regional dialects and accents are not taken into account; go by the standard pronunciation.
• Are the goods and services the same or related? If a mark that sounds the same is used in an unrelated field, the risk usually drops.

The cheapest time to shorten the list of candidates is when you choose the name. We explain how to come up with a name that carries low phonetic risk from the start in our guide on how to choose a brand name. For marks derived from a surname, spellings with and without Turkish characters can appear as separate records; we cover questions specific to personal names in our article on registering your own name as a trademark.

A Checklist for Phonetic Searches

The following list helps you screen pronunciation variants without missing any and document the result:

1. Read your mark aloud and have two different people write it down as they hear it; note the spellings they produce.
2. Search the versions with and without Turkish characters separately.
3. Try the k/c/q, ş/s/sh, ç/c/ch, w/v, x/ks and j/c equivalents one by one.
4. Search forms with doubled letters, dropped letters and added suffixes.
5. Try joined, separate and hyphenated spellings.
6. If the mark contains a numeral or symbol, search it both in figures and in words.
7. Search a foreign word under spellings that reflect its Turkish reading and its original reading.
8. Widen the list with root and wildcard searches, then narrow it with a class filter.
9. Record the date of the search, the variants you tried and what you found.

Skipping variants is one of the most common mistakes in trademark searching; we list the others in our article on trademark search mistakes.

Summary: Key Points on Phonetic Similarity

• Databases match letters; to find marks that sound the same, you have to search the variants yourself.
• Phonetic similarity carries different weight in the two channels: in the Office's ex officio examination, aural similarity alone is generally not enough, while on opposition it is weighed as a separate dimension.
• Letter equivalents, suffixes and spelling format, numerals and symbols, foreign pronunciation and plural forms are the five main groups of variants.
• Root and wildcard searches cover the variant list with fewer queries.
• Prioritize results by opening syllable, syllable structure and the relationship between goods and services.

Shall we screen the pronunciation variants of your mark together?

Send us your trademark and a description of your business through our contact page; our team, which works with trademark attorneys authorized before TÜRKPATENT, will run the pronunciation-variant, root and wildcard searches class by class and report the phonetic risk with the reasons behind it.